A thermomechanical perspective on caldera …models to Santorini Volcano in Greece where recent...

A thermomechanical perspective on caldera formation and classification

Shanaka L de Silva, Patricia M Gregg

Oregon State University, U.S.A

E-mail: [email protected]

Decades of research have led to a mature understanding of caldera formation and evolution. In particular

a temporal model for caldera formation (Smith and Bailey, 1968) is now well understood and morphological

classifications based on geological studies (e.g. Lipman, 1997; Cole et al, 2005), recently informed by analogue

models (Acocella, 2007), capture the final surface manifestations and styles of collapse and inferred relationships to

chamber depth, size, and shapes (Lipman, 2000; Roche et al., 2000; Kennedy, 2000). However, our understanding

of the processes that lead to these morphologies and the temporal evolution remain murky. In particular, what

initiation and triggers catastrophic caldera collapse is unclear. The general two stage model of collapse (Stages 2

and 3 of Smith and Bailey, 1968) is the paradigm that prevails (Druitt and Sparks, 1984; Kennedy and Stix, 2003).

In this model an initial eruption is triggered by overpressure in the magma chamber. Evacuation of magma and

bleeding off the overpressure leads to an underpressured condition in the magma reservoir that triggers caldera

collapse and pyroclastic flow generation. However, it has been increasingly recognized that many of the largest

calderas do not conform to this two-stage model and involve catastrophic caldera collapse at the inception of

eruption (Christiansen, 2005; de Silva et al., 2006). Moreover, many of these large calderas exhibit a significant

structural resurgence that smaller systems do not. Large caldera collapses result in collapse of the entire upper

crust above the magma system, smaller calderas form within the edifice. The details of the physical volcanology of

the eruptions and the resulting ignimbrites are quite different.

These differences between large and small calderas can be understood from a thermomechancal perspective.

Generation of viable large magma bodies requires elevated thermal fluxes and thermally mature crust (de Silva and

Gosnold, 2007; Annen, 2009) that plays a fundamental role in the rheology of magma/host rock interface (Jellinek

and DePaolo, 2003) and ultimately controls magma reservoir growth, pressure evolution, surface uplift, faulting,

and caldera size (Gregg et al, 2012, 2013). In particular thermomechanical considerations suggest that in large

systems the catastrophic caldera eruptions are "externally" triggered by mechanical failure of the roof. Smaller

systems are "internally" triggered by overpressure within the magma reservoir. These different modes of triggering

largely control the final form of the calderas and the character of the eruption and associated pyroclastic deposits.

We suggest that catastrophic calderas can be fundamentally divided into two size divisions with an approximate

boundary at 10km diameter, 100 km3 and VEI=7. Ultimately this is a thermal or thermomechanical divide.

IAVCEI 2013 Scientific Assembly - July 20 - 24, Kagoshima, JapanForecasting Volcanic Activity - Reading and translating the messages of nature for society

©IAVCEI 2013 Scientific Assembly. All rights reserved. 781

3A1_3C-O1 Room A5 Date/Time: July 23 8:45-9:00

Fail or not to fail? Reservoir mechanics and ground deformation before large magnitudeeruptions of intermediate magmas

Joachim H Gottsmann, Karen Strehlow

School of Earth Sciences, University of Bristol, UK


The current knowledge base on ground deformation prior to the eruption of intermediate magma is skewed

towards small magnitude eruptions. The long-term (order of years) pre-eruptive ground deformation of a future

large magnitude intermediate eruption similar to the caldera-forming Great Tambora eruption in 1815 is thus poorly

understood.

Here, we explore a potential long-term pre-eruptive geodetic signature of a growing magma reservoir of

intermediate composition using numerical mechanical modelling.

We start by exploiting cyclic ground deformation data from the currently active Soufrière Hills volcano to

constrain the pre-eruptive subsurface stress history of a comparably small magmatic system of intermediate

composition. We first define a reservoir failure criterion based on rock tensile strength under the simple assumption

of elastic mechanical behaviour of surrounding rocks. Given observed pre-eruptive deformation amplitudes and

petrologically deduced storage conditions, the results indicate that invoking mechanical elasticity is problematic to

explain cyclic eruptive behaviour at SHV over the past decade. Time-dependent stress relaxation must play a first

order role.

Up-scaling to match magmatic conditions prior to the Great Tambora eruption demonstrates a first order

influence of edifice load, topography, and mechanical heterogeneity of encasing rocks on the stress distribution

and the resultant deformation field upon reservoir growth. Again time-dependent stress relaxation is found to

control growth of an intermediate magma reservoir of substantial size and thereby inhibiting pre-mature failure

upon recharge.

The modeling also shows that a dynamic failure criterion is more plausible compared to a static criterion

based on rock tensile strength to characterize reservoir failure at lithostatic pressures of between 100 and 220

MPa, i.e., within a pressure window constrained petrologically for many recently erupted intermediate magmas.

Combining the numerical results with published thermal and petrological constraints on reservoir evolution, we

deduce a limiting permissible volumetric strain rate upon reservoir recharge of 108 s-1. Growth of a large body

of eruptible magma of intermediate composition may be favoured at strain rates <108 s-1, whereas pre-mature

reservoir failure may occur at strain rates >108 s-1. The associated long-term pre-eruptive ground uplift of the

pre-1815 Tambora volcano by the incremental growth of a relatively short-lived (compared to silicic) intermediate

magma reservoir is predicted at well below 1 cm/year. The results may be useful for assessing the long-term

pre-eruptive signs of future large magnitude eruptions




Thermal-mechanical models of magma chamber pressurization: Implications forchamber growth and caldera formation

Patricia M Gregg1, Shanaka L de Silva1, Eric B Grosfils2

1Oregon State University, United States, 2Pomona College, United States


Rapid magma chamber overpressurization due to the influx of new material or the exsolution of volatiles is often

cited as a triggering mechanism for the eruption of a shallow magma chamber. Specifically, the increase in

chamber overpressure may drive the tensile stress along the chamber boundary above the tensile strength of

the surrounding wall rock and trigger a dike or sill intrusion and/or eruption. A major limitation to previous numerical

and analytical models of magma chamber rupture is that overpressure is assumed constant and applied as a

fixed boundary condition. This fixed boundary condition does not allow magma chamber expansion to dissipate

overpressure. Understanding how overpressures build is critical for determining eruption triggers. In this study, we

utilize thermomechanical models to investigate how magma chambers overpressurize as the result of either the

influx of new melt or volatile exsolution. By incorporating an adaptive reservoir boundary condition we are able

to track how overpressure dissipates as the magma chamber expands to accommodate internal volume changes.

We find that the size of the reservoir greatly impacts the resultant magma chamber overpressure. In particular,

overpressure estimates for small to moderate sized reservoirs (1-10 km3) are up to 70% lower than previous

predictions. Furthermore, our models indicate that systems >100 km3 do not generate large overpressures,

because magma volume changes are accommodated by deformation of the viscoelastic host rock. We apply our

models to Santorini Volcano in Greece where recent seismic activity and ground deformation observations suggest

the potential for a large caldera forming eruption. A viscoelastic model with a fixed overpressure boundary condition

predicts a much lower magma flux than calculated using a Mogi source. This suggests that the increase in volume

flux is much more modest than has been indicated in previous investigations. Conversely, the incorporation of an

adaptive boundary condition reproduces Mogi estimations of a high magma flux and suggests that the magma

reservoir present at Santorini may be quite large. Model results further indicate that, if the magma chamber is >100

km3, overpressures generated due to the high magma flux may be below the tensile strength of the host rock, thus

requiring an additional triggering mechanism for eruption.




Caldera development during explosive eruptions

John Stix

McGill University, Canada


An interesting question is the likelihood that an explosive eruption results in caldera formation. Large eruptions

emitting 102-103 km3 of magma invariably form substantial surface subsidence structures, but smaller eruptions

emitting 5-10 km3 may not. These smaller events show variable behaviour, which is generally unpredictable,

ranging from well-developed caldera structures (e.g., Pinatubo) to no calderas at all (e.g., Huaynaputina). For such

events, is it possible to establish, during or even prior to the eruption, the probability that a caldera will form? Such

an assessment is important, as the style of the eruption depends in part upon whether a caldera forms. One clear

control is depth to the top of the magma reservoir; as depth increases, surface subsidence is reduced due to the

crustal column between the surface and the top of the reservoir. A second important consideration is the amount

of eruptable magma in the reservoir. An eruption that efficiently drains such eruptable magma from the reservoir

may suddenly cease, putting a halt to caldera development. The caldera volume is commonly smaller than the

volume of erupted magma, and this is a reflection of the mismatch between efficient magma extraction and surface

caldera response. There is general recognition that many volcanic systems are underlain by a series of magma

reservoirs stacked vertically in the crust and connected at times. There may thus be a balance between the relative

roles of a shallow reservoir at 4-10 km depth and a second reservoir at deeper crustal levels which together may

influence surface conditions including caldera subsidence. If the shallow chamber is small or largely crystallized,

then the deeper reservoir may play a greater role in controlling the nature and size of the eruption. If the shallow

chamber is large or largely liquid, then its influence will dominate. A fundamentally important aspect of this stacked

magmatic system is that the shallow and deeper reservoirs are connected, allowing magma to be transferred from

the deep to the shallow reservoir. The connection may be established, before, during, and/or after the eruption.

Large explosive eruptions commonly appear to have this connection established prior to the eruption, allowing deep

magma to (a) interact and mobilize shallow magma and (b) trigger the eruption. Magma erupted from the shallow

reservoir unloads the system, providing a feedback for inflow of magma from below, perhaps at accelerating rates

at later stages of the eruption and immediately afterward. Forecasting caldera formation during an eruption may be

possible. Since surface caldera development is commonly a late-stage eruptive phenomenon which is preceded by

subsurface faulting and subsidence, geophysical approaches (e.g., seismology, geodesy, gravity) may be used to

identify subsurface processes such as faulting and development of voidspace which migrate upward in time.




Analogue caldera collapse models characterized with radiography and computerizedX-ray micro-tomography

Sam Poppe1, Eoghan P. Holohan2, Matthieu N. Boone3, Elin Pauwels3, Veerle Cnudde4, MatthieuKervyn1

1Department of Geography, Earth System Sciences, Vrije Universiteit Brussel (VUB), Belgium, 2GFZPotsdam, Section 2.1, Germany, 3Centre for X-ray Tomography (UGCT), Department of Physics andAstronomy, Ghent University, Belgium, 4UGCT, Department of Geology and Soil Science, Faculty of

Science, Ghent University, Belgium


Analogue models are widely used to investigate volcano-tectonic processes, their structural geometries, kinematics

and dynamics. Past models explored caldera collapse structures mainly via 2D model cross–sections. For

interpreting field and monitoring data, however, it is essential to understand the kinematics and geometry of caldera

collapse structures in 3D. We applied high resolution radiography and computerized X–ray micro–tomography (µCT)

to image the deformation during analogue fluid withdrawal in small-scale caldera collapse models. High resolution

interval radiograph sequences provide an unprecedented ′2.5D′ documentation of the surface and internal model

geometries, as well as of the kinematics of a collapsing column into an emptying fluid body. Subsidence was

controlled by ring faults and associated with dilatation of the analogue granular material within the collapsing

column. The subsidence rate within the collapse column showed three main phases: 1) Upward migration of a

high velocity zone associated with ring fault propagation, 2) Rapid subsidence with the highest subsidence rates

within the uppermost subsiding volume, 3) Relatively slower subsidence rates over the whole column but with

intermittent accelerations of discrete sections of the column. By using radiograph sequences, it is possible to

obtain a continuous observation of fault propagation, down sag mechanisms and the subsequent development

of collapse structures in a non-destructive manner. µCT scans of the post collapse model enable a full 3D

reconstruction of the model and its internal structure. The models highlight the possibilities and limitations of µCT

scanning to qualitatively image and quantitatively analyse deformation of analogue volcano–tectonic experiments.

Despite some practical model limitations, the radiograph and µCT method is hence a step towards a quantitative

documentation of analogue models that would render experimental data more immediately comparable to recent

monitoring data. The models also carry the potential for a better understanding of the kinematics of caldera collapse

amongst a variety of volcano–tectonic processes.




Coupled long- and short-term eruption precursors during caldera unrest

Robert M Robertson

AON Benfield UCL Hazard Research Centre, University College, London, England


Robert Robertson and Christopher Kilburn

Aon Benfield UCL Hazard Centre, Department of Earth Sciences, University College London, Gower St,

London WC1E 6BT, UK.

Short-term forecasts rely on accelerations in precursory signals. Although the strategy has proved suitable

at stratovolcanoes reawakening after centuries of repose, it has been less successful at large calderas, where

unrest may continue for decades or more. Here we argue that patterns of short- and long-term unrest are strongly

coupled at large calderas, so that reliable interpretations of short-term precursors require knowledge of preceding

episodes of long-term unrest.

During unrest, potential precursors may show significant variations in their rates of change with time. More

than one episode of increased rate may decay without eruption. Associated warnings of eruption can be viewed as

false alarms by local communities, who may then be reluctant to respond during future alarms. A classic example

is the 21 years of unrest before the 1994 eruption in Papua New Guinea. An uplift of c. 2.3 m was recorded

across the north of the caldera. Nearly 40% of the uplift occurred in 1982-84, accompanied by elevated rates of

volcano-tectonic (VT) seismicity. The uplift rate of c. 0.45 m/yr was a magnitude larger than typical values over

the 21-year interval. The 1982-84 crisis ended without an eruption. When an eruption finally occurred in 1994, it

began without a significant increase in uplift rate and with an elevated VT event rate only 24 hours beforehand. An

apparent contradiction thus exists in which significant precursory changes occurred without eruption, whereas an

eruption occurred without significant precursory changes.

The contradiction is resolved by comparing changes of precursors with each other, rather than with time.

The results yield an exponential dependence of VT event number on uplift. The trend is consistent with a single

precursory sequence, during which damage accumulates in the crust until a new pathway is formed to feed an

eruption. The amount of damage increases as the crust is stretched under a combination of (1) a build-up of

pressure in an underlying magma reservoir and (2) movement of magma through the crust. Thus, the 1973-94

sequence can be related to a long-term increase in magma pressure, coupled with the shallow intrusion of a small

batch of magma during 1982-84. At the time of intrusion, the amount of accumulated damage was too small to

propagate a new pathway to the surface and so the event culminated in a non-eruptive crisis. By 1994, however,

the amount of damage accumulated since the early 1970s had reached a critical amount, such that another

episode of magma movement triggered the opening of a new pathway and led to an eruption after a very short

interval of elevated precursory signals.




Growth and rupture of the Minoan magma body, Santorini

Timothy H Druitt

Lab Magmas et Volcans, Blaise Pascal University, Clermont-Ferrand, France


The eruption of large silicic magma chambers following prolonged periods of crustal intrusion raises the question

of what ultimately initiates venting. It is commonly invoked that the chamber pressure and or crustal stress

state increases until some threshold is exceeded. I present evidence bearing on the trigger of the late 17th

century BC Minoan eruption of Santorini, which discharged 30-60 km3 of magma. Three juvenile components

were erupted: rhyodacitic pumices (>99 %), cauliform andesitic enclaves and crystal-rich andesitic pumices. The

rhyodacite has 70-71 wt% of silica with about 10 wt% of phenocrysts set in rhyolitic glass. The enclaves have

cauliform surfaces, variably ellipsoidal to tabular shapes, silica contents of 52-61 wt% and consist of 20-40 wt% of

phenocrysts and fragments of disrupted gabbro set in silicic andesitic groundmass. Some have skins of adhering

pumice showing that they are quenched mafic enclaves released from the rhyodacitic host on eruption. The tabular

enclaves may be detached selvages from the margins of a composite dyke. The crystal-rich pumices have silica

contents of 52-64 wt%, 40-60 wt% of crystals, and rhyolitic glass. They grade texturally into glass-bearing and

holocrystalline nodules of hornblende diorite, and are interpreted as the contents of a variably crystallized intrusion

- melt-dominated interior, crystal-dominated margins and holocrystalline carapace - that were disrupted, mixed

together and discharged during the eruption. The enclaves, crystal-rich pumices and dioritic nodules formed from

a batch of Ba-rich, Zr-poor andesite that is unique in Santorini products of the last 550 ky. Their high Ba/Zr

ratios shows that the andesites are not related in any simple way to the rhyodacite. The crystal-rich pumice is the

dominant magmatic component at the base of the eruption sequence. It appears that the rhyodacite made its way

to the surface by exploiting the pre-existing andesitic/dioritic intrusion, pushing the crystal-rich contents ahead of it

and entraining enclaves from the less crystalline interior. A recent study of crystal zoning patterns in the rhyodacite

concluded that the Minoan magma reservoir experienced a spurt of high-flux growth during the century to months

prior to eruption onset, due to recharge by large volumes of principally silicic magma. However, eruption finally

occurred only once the rhyodacite encountered a suitable pathway to the surface, provided by the mushy Ba-rich

intrusion. Either (1) a dyke of andesite feeding the Ba-rich intrusion intersected the main magma chamber, or

(2) the main magma chamber inflated under the influx of silicic melt until it intersected the Ba-rich intrusion. The

andesites lubricated ascent of the rhyodacite, leading to eruption.




The other side of caldera unrest: why do caldera subside?

Micol Todesco, Francesca Quareni, Antonio Costa

Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Bologna, Italy


Calderas are known for their restless behavior: remarkable ground deformation is

commonly observed during non-eruptive periods, sometime accompanied by seismicity

and by significant changes in the chemistry and flow rate of discharged fluids. Uplift

phases usually draw attention, as they may be due to shallow magma intrusion, and could

therefore prelude to eruptive activity. However, the peculiar feature that characterizes

calderas since their very formation is subsidence. Generated by a syn-eruptive collapse of

the magma chamber roof, calderas commonly feature a long-term, secular subsidence,

that is usually ascribed to magma cooling and contraction and/or to compaction of the soft

volcanic deposits. Over a shorter time scale, subsidence may follow episodes of volcanic

unrest, as a result of either magma movement or due to the discharge of volcanic gases.

Despite its common occurrence and its relevance in the caldera evolution, a systematic

assessment of the processes that could drive and explain caldera deflation is still missing.

In this work we focus on the secular evolution of the Campi Flegrei caldera. Thanks to the

presence of urban settlements along the sea shore, dating back to Roman times, this

caldera is a unique site to evaluate the time scales associated with uplift and subsidence

phases. Based on this formidable benchmark, we review the geological processes that

may be responsible for subsidence, searching for time scales and magnitudes of

displacement that are consistent with the available constraints. We believe that a

comprehensive analysis that considers all the aspects of ground motion could provide a

better image of the deep phenomena that fuel the caldera restless activity.




Caldera resurgence during magma replenishment and rejuvenation at Valles and LakeCity calderas

Ben M Kennedy1, John Stix2, Jack Wilcock2

1University of Canterbury, New Zealand, 2McGill Uiversity, Canada


A key question in volcanology is the driving mechanisms of resurgence at active, recently active, and ancient

calderas. Valles caldera in New Mexico and Lake City caldera in Colorado are well-studied resurgent structures

which provide three crucial clues for understanding the resurgence process. (1) Within the limits of 40Ar/39Ar dating

techniques, resurgence and hydrothermal alteration at both calderas occurred very quickly after the caldera-forming

eruptions (tens of thousands of years or less). (2) Immediately before and during resurgence, dacite magma

was intruded and/or erupted into each system; this magma is chemically distinct from rhyolite magma which was

resident in each system. (3) At least 1 km of structural uplift occurred along regional and subsidence faults which

was closely associated with shallow intrusions or lava domes of dacite magma. These observations demonstrate

that resurgence at these two volcanoes is temporally linked to caldera subsidence, with the upward migration of

dacite magma as the driver of resurgence. Recharge of dacite magma occurs as a response to loss of lithostatic

load during the caldera-forming eruption. Flow of dacite into the shallow magmatic system is facilitated by regional

fault systems which provide pathways for magma ascent. Once the dacite enters the system, it is able to heat,

remobilize, and mingle with residual crystal-rich rhyolite remaining in the shallow magma chamber. Dacite and

remobilized rhyolite rise buoyantly to form laccoliths by lifting the chamber roof and producing surface resurgent

uplift. The resurgent deformation caused by magma ascent fractures the chamber roof, increasing its structural

permeability and allowing both rhyolite and dacite magmas to intrude and/or erupt together. This sequence of

events also promotes the development of magmatic-hydrothermal systems and ore deposits. Injection of dacite

magma into the shallow rhyolite magma chamber provides a source of heat and magmatic volatiles, while resurgent

deformation and fracturing increase the permeability of the system. These changes allow magmatic volatiles to rise

and meteoric fluids to percolate downward, favoring the development of hydrothermal convection cells which are

driven by hot magma. The end result is a vigorous hydrothermal system which is driven by magma recharge.




Caldera unrest at Taupo Volcano, New Zealand: intensity, impacts and mitigation

Sally H Potter1, Gill E Jolly1, Vince E Neall2, David M Johnston3

1GNS Science, 114 Karetoto Road, Wairakei, New Zealand, 2Massey University, Private Bag 11 222,Palmerston North, New Zealand, 3Joint Centre for Disaster Research, Massey University, P.O. Box 756,

Wellington, New Zealand


Taupo Volcano is considered one of the most active caldera centres in the world. Its last eruption occurred in

232 AD, in a 35 km³ volume (DRE), rhyolitic, caldera-forming eruption which devastated a large portion of the

Central North Island of New Zealand. Prior to eruptions volcanoes exhibit signs of unrest commonly in the form

of seismicity, ground deformation or hydrothermal or geochemical changes. The integration and interpretation

of these phenomena is considered key to eruption forecasting. The majority of unrest episodes at calderas do

not result in eruptions leading to potential "false alarms". Identifying the range of possible future multi-parameter

unrest activity based on previous events at the volcano contributes to eruption forecasting.

This research involves a historical chronology of all volcanic activity observed at Taupo Caldera since

written records began in the 1850s. Document analysis of newspaper articles, local and scientific literature,

correspondence and analysis of monitoring data contributed to the dataset. From this, a catalogue was created

summarising 91 episodes of heightened activity.

The wide range of intensity of episodes found in this research questions the definitions of the terms "background"

and "unrest" commonly used in the literature. The complete range of activity at any type of volcano seen during

periods of quiescence (including caldera unrest) is described here as Non-Eruptive Volcanic Activity (NEVA).

The physical impacts of NEVA at Taupo Caldera include subsidence of 3.7 m over a period of months, seiches in

the lake which fills the caldera and episodes with an average of 100 earthquakes felt per day. Societal impacts

reported include anxiety, self-evacuations, damage to infrastructure and a perceived impact on the local and

regional economies.

In order to help mitigate the impacts of future caldera unrest in New Zealand, a Caldera Advisory Group

has been formed. This is a multi-agency strategic planning group organised by regional councils. Other members

include the Ministry of Civil Defence and Emergency Management, local councils and GNS Science. Outputs

of this Group include a caldera unrest information sourcebook for non-scientists, and a scenario for planning

purposes based on the past NEVA at Taupo Caldera.




The Oskjuvatn caldera North Iceland revealed by detailed bathymetric study

Armann Hoskuldsson, Thor Thordarson

University of Iceland, NORVOL, Institute of Earth sciences, Iceland


The Dyngjufjoll volcanic complex has an area of some 400 km2 with a maximum elevation of 1,516 m a.s.l., rising

600 to 800 m above the surrounding highland plateau. The oldest exposed parts of the centre, dominantly basaltic

in composition, were formed in subglacial and subaquatic eruptions and deposited as hyaloclastites forming the

elevated topography. The volcanic centre is surrounded and partially covered by extensive Holocene lava flows

erupted from fissures striking NE-SW. There are at least 4 if not 5 calderas that have developed within the volcanic

complex during the past 700 ka. By far the largest and most prominent one is the 45-km2 Askja caldera centrally

located within the Dyngjufjoll massive, thought to have formed some 10 ka ago. The floor of the caldera is covered

with Holocene lavas which issued from fissures mostly along the caldera rims. The youngest caldera is the nested

Oskjuvatn caldera formed in the southeast corner of the main Askja caldera during and after the 1875 A.D. eruption.

The Oskujvatn caldera is a lake-filled caldera, which has hampered its geological observation up to now. The

caldera is 5 km in diameter (area, 18 km2). The maximum depth of the caldera lake is 205 m in the western half of

the caldera. Rims of the caldera rise >60 m above the lake surface, indicating a total depth of no less than 260 m

for the structure. Maps made by a Danish expedition in 1876 show that location of the deepest part of the caldera

has changed. We also observe several eruptive vents that are unaccounted for, both along the main ring fault and

within the eastern half of the caldera. Several large geothermal areas are observed on the bottom of the caldera,

one in the west below the Myvetningahraun lava flow and the other in the east in relation to, until know, unknown

volcanic vent. Analysis of historical accounts shows that the Oskjuvatn caldera was not fully developed until 1932

(Hartley and Thordarson, 2012), while internal unconformities in the 28-29 March 1875 tephra deposit indicate that

the initiation of the collapse coincides with onset of the eruption. This suggests that the formation of the Oskjuvatn

caldera took more than 50 years. These observations along with the new bathymetric map of the Oskjuvatn caldera

will be presented and discussed.




Investigating the state of the reservoir below the Rabaul caldera (Papua New Guinea)

Caroline Bouvet de Maisonneuve1, Fidel Costa1, Christian Huber2, Herman Patia3

1Earth Observatory of Singapore, Nanyang Technological University, Singapore, 2School of Earth andAtmospheric Sciences, Georgia Tech., USA, 3Rabaul Volcano Observatory, Papua New Guinea


The area of Rabaul (Papua New Guinea) consists of at least seven - possibly nine - nested-calderas that have

formed over the past 200 ky. The last caldera-forming eruption occurred 1400 y BP, and produced about 10

km3 of crystal-poor, two-pyroxene dacite. The most recent major explosive activity in the area was preceded by

decade-long unrest (starting in 1971) until the simultaneous eruption of Vulcan and Tavurvur, two vents on opposite

sides of the caldera (September 19th, 1994). Ten years prior to the eruption, a peak of seismic activity and

deformation occurred, which promoted part of the local population to evacuate spontaneously. The recent history

at Rabaul is therefore a perfect example of the complexity and problems associated with volcanic unrest and activity

in a caldera setting. In addition, it is possible to study the deposits of the eruptions that ensued from the prolonged

period of seismic and deformational activity, and therefore gain insights into the processes responsible for the

unrest. The 1994-to-present deposits show clear signs of mixing/mingling between basaltic and dacitic magmas,

in the form of banded pumices, quenched mafic enclaves, and hybrid bulk rock compositions. We estimate that at

least 20-35 wt% basalt has mixed with the resident silicic magma. The time scales of mixing indicated by kinetic

modeling of plagioclase zoning suggest that replenishment events coincide with the main period of unrest (1971

to 1985). We use a petrological, geochemical, and numerical modeling approach to investigate the current state

of the main reservoir and its evolution since the last caldera-forming eruption. Our working model is that basaltic

melts ascend to shallow crustal levels before intruding a main silicic reservoir beneath the Rabaul caldera. Storage

depths and temperatures estimated from mineral-melt equilibria and rock densities suggest that basalts ascend

from ca. 20 km (600 MPa) to ca.7 km (200 MPa) and cool from ca. 1150-1100 C before intruding a dacitic magma

reservoir at ca.950 C. Depending on the state of the reservoir and the volumes of basalt injected, the replenishing

magma may either trigger an eruption or cool and crystallize. Preliminary geochemical modeling suggests that the

caldera-forming dacitic magma could be generated by fractional crystallization of the basaltic magma at shallow

depths (ca.7 km, 200 MPa) and under relatively dry conditions (less than 3 wt% H2O). We use evidence from major

and trace element geochemistry, volatile contents, and the comparison of successive eruptions since 1400 y BP to

address the question of whether another potentially caldera-forming magma is presently brewing beneath Rabaul.




Magmatic-tectonic triggers leading to the eruption of multiple isolated magma batchesand twin caldera collapse in the Taupo Volcanic Zone, New Zealand

Florence Begue1, Darren M Gravley1, Chad D Deering2, Ben M Kennedy1, Isabelle S Chambefort3

1University of Canterbury, New Zealand, 2University of Wisconsin-Oshkosh, USA, 3GNS Science, NewZealand


Calderas, their associated structures, and the geometry of their pre-collapse magma reservoir(s) have been the

subject of much scientific intrigue, largely because they represent the largest scale of volcanic eruptions and our

knowledge has only been gleaned from ancient examples. However, detailed field, analogue, and petrochemical

studies of several calderas reveal an often complex arrangement of magma bodies that are chemically distinct

from each other. In addition, the arrangement and geometry of these magma bodies can be dictated by regional

tectonic structures that are activated during eruption and collapse. Here, we present new petrochemical evidence

for a complex subvolcanic plumbing system including multiple juxtaposed magma bodies that erupted over a

remarkably short timeframe to form twin calderas 30 kilometers apart.

The Rotorua and Ohakuri calderas in the central Taupo Volcanic Zone (TVZ) formed 240 ka with the

eruption of >245 km3 rhyolitic magma associated with the Mamaku and Ohakuri ignimbrites, respectively. As a

consequence, a large volcano-tectonic depression spanning the area between the two calderas formed. New

pumice glass and melt inclusion data (combined with the existing bulk-rock and mineral chemistry) supports

a petrogenetic model for six different magma batches extracted from the same source region, i.e. a laterally

extensive and continuous intermediate mush zone beneath the Rotorua and Ohakuri calderas. Minor geochemical

differences in the batches, including fluid mobile elements U, Cs, and Li and volatiles (CO2 and H2O), suggest

different extraction conditions for the emplacement of the magma batches. However, combining the chemistry with

the timing of eruption for each batch (recorded in the deposit stratigraphy) suggests a more tantalising picture

for the spatial distribution of the magma batches in the upper crust and a close interplay between tectonics,

magmatism and volcanism. In particular, the initial plinian airfall eruption from Ohakuri erupted simultaneously

with the Mamaku ignimbrite from Rotorua and they share a similar trend in Cs and Li concentrations, and CO2

values that suggest, if volatile saturation was reached, melt inclusion entrapment at pressures ranging from 75

to 150 MPa. The chemistry suggests that the magma batches for the earliest eruptions may have been fed by

a connected rhyolitic dyke on the western margin of the large volcano-tectonic depression. An elongated rhyolite

dome with the same 240 ka age and same chemistry also lies on the western margin of the depression between

the two calderas; further supporting the idea of a laterally extensive dyke. The evacuation of these magmatically

linked melt batches via a laterally extensive dyke could have initiated rupture of regionally linked faults enough to

trigger the eruption of the chemically distinct magma batches associated with the Ohakuri ignimbrite and collateral

subsidence of the area between the two calderas.




First recorded eruption of Nabro volcano, Eritrea, 2011

Berhe Goitom1, Ghebrebrhan Ogubazghi1, Jon Blundy2, Simon Carn3, Amy Donovan4, David Fee5,Raphael Grandin6, Jo Hamlyn7, James Hammond8, Mike Kendall2, Clive Oppenheimer4, Tim Wright7,

Talfan Barnie4, Bruno Scaillet9, Carolina Pagli7

1Eritrea Institute of Technology, Eritrea, 2University of Bristol, UK, 3MTU, USA, 4University ofCambridge, UK, 5UAF, USA, 6IPGP, France, 7University of Leeds, UK, 8Imperial College, UK, 9ISTO,

France


The 2011 eruption of Nabro began with few detected precursory signals. At the time there was no seismic network

operating in Eritrea. Nevertheless a rapid response on the ground led to timely evacuation of settlements within

the two calderas and on the flanks of the volcano. Several thousand people were displaced and cared for in

temporary camps in the region. In broad terms, Nabro is sited in the extensional zone of Afar, close to the

Mesozoic crustal block of the Danakil Alps. It reaches a maximum elevation of over 2200 m above sea level, and

has an 8 km wide summit caldera and associated ignimbrites. The 2011 eruption began on 12 June following

intense seismicity. It is the first eruption of Nabro on record, highlighting the potential of caldera systems to

erupt with limited warning. It is also the first seismicity of note instrumentally recorded in this part of the rift.

Remarkably, the Nabro plume significantly perturbed stratospheric aerosol optical depth reflecting a sizeable SO2

emission (of order 1.5 Tg). We present here a preliminary synthesis of the nature and causes of the eruption

based on multiple observations (satellite remote sensing, seismology, infrasound records, ground observations,

and petrological characterisations).




Episodes of magmatic fluid injection into hydrothermal systems and caldera unrests.The case of Campi Flegrei (Italy)

Giovanni Chiodini1, Stefano Caliro1, Carlo Cardellini2, Prospero De Martino1, Annarita Mangiacapra1,Zaccaria Petrillo1

1Istituto Nazionale di Geofisica e Vulcanologia, Napoli, OV, Italy, 2Universita’ di Perugia, Dipartimento diScienze della Terra, Italy


The isotopic signatures of many fumaroles from volcanoes, and in particular from calderas, indicate that the feeding

hydrothermal systems are recharged by a mixture of shallow waters, of meteoric or marine origin, with magmatic

fluids. In the few cases where long time series of chemical and isotopic compositions of the fumarolic vents

are available, the data suggest that the mixing processes occur during short periods during which large amount of

magmatic fluids are injected in the hydrothermal system. In this frame Campi Flegrei is the best studied case. Since

1980, the fumarole compositions and their variation in the time are known together with the seismic activity and the

ground deformation of the caldera. On the base of an accurate analysis of the long geochemical time series, i.e.

variations in the main component of the fumaroles as well in the minor gas species and in their isotopic composition,

we recognised that thirteen episodes of injection of magmatic fluids into the hydrothermal system occurred from

1980 to 2012. The sudden arrive of the magmatic fluids causes the pressurization of the hydrothermal system

and in turn pulsed ground uplift episodes and earthquakes, a process which well explain the correlation in the

time among geochemical and geophysical signals. The process was simulated with a physical numerical model

able to treat energy and fluid dynamics in a hydrothermal system characterised by bi-phase (vapor and liquid) and

bi-component (water and carbon dioxide) fluids. The model was constrained by the flux of CO2 and of thermal

energy measured at the surface as well as by the rock properties of the volcanic products filling the caldera. The

results indicate that each episode of magma degassing involve an amount of fluids of the same order of magnitude

as that involved in medium-small size eruptions. Furthermore the cumulative curve of the masses injected in each

episode shows an inversion in 2000 which divides a preliminary period of decrease in the flux of magmatic fluids

from the present phase of increasing activity, suggesting the beginning of a new unrest phase at Campi Flegrei. The

approach used here is potentially a powerful tool for quantifying the evolution of hydrothermal systems undergoing

magmatic fluid injections. This opens the door to accurately detecting and correctly interpreting early signals of

volcanic unrest in potentially hazardous caldera settings.



3P1_3C-O15 Room A5 Date/Time: July 23 15:00-15:15

Mapping long-term vent opening in a caldera setting with uncertainty estimation:application to Campi Flegrei caldera (Italy)

Andrea Bevilacqua1, Roberto Isaia3, Antonella Bertagnini2, Marina Bisson2, Celine Fourmentraux2,Franco Flandoli4, Enrico Iannuzzi5, Augusto Neri2, Mauro Rosi5, Carlo Scirocco6, Stefano Vitale6

1Scuola Normale Superiore, Pisa, Italy, 2Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Pisa,Pisa, Italy, 3Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Vesuviano, Napoli, Italy,

4Universita’ di Pisa, Dip.to di Matematica, Pisa, Italy, 5Universita’ di Pisa, Dip.to di Scienze della Terra,Pisa, Italy, 6Univ. di Napoli Federico II, Dip.to di Scienze della Terra, dell’Ambiente e delle Risorse,

Napoli, Italy


Campi Flegrei is an active volcanic area located in the Campanian Plain, along the Tyrrhenian margin of the

southern Apennines (Italy), dominated by the formation of a 12 km large, resurgent caldera. The great majority

of the eruptions have been explosive, variable in magnitude and intensity and characterized by the generation of

remarkable ash fallout and pyroclastic density currents deposits. In this study we present results of field work

and statistical analysis of past eruptive activity aimed at producing long-term probabilistic maps of vent opening at

Campi Flegrei. Field work was focused on the structural and morphological nature of the caldera and particularly on

the reconstruction of the location of past eruptive vents as well as of main faults and eruptive fissures formed in the

last 15 kyr of activity. The statistical analysis performed accounted for the spatial distribution of past vent locations

but was flexible enough to incorporate the heterogeneous geological information available, such as the density of

faults/fissures or the clue of possible past vents hidden by the more recent activity. One key objective of the analysis

was to directly incorporate into the maps the main uncertainties affecting the system. This was done by adopting

appropriate density distributions of the probability of vent opening of the different areas of the caldera and by relying

on expert judgement. Results allowed to quantify the influence of the different theoretical assumptions and sources

of uncertainty on the long-term mapping of vent opening. The distributions obtained represent the starting point for

the production of long-term ash fallout and pyroclastic density hazard maps at Campi Flegrei caldera.




The Campi Flegrei Deep Drilling Project: using borehole data to model caldera unrest

Stefano Carlino1, Anna Tramelli1, Christopher Kilburn2, Monica Piochi1, Renato Somma1, ClaudiaTroise1, Giuseppe De Natale1

1Istituto Nazionale di Geofisica e Vulcanologia - Osservatorio Vesuviano, Italy, 2Aon Benfield UCLHazard Centre, Dept of Earth Sciences, UK


Campi Flegrei is a densely populated, active caldera immediately west of Naples in southern Italy. It has a

well-documented history of unrest since Roman times, involving caldera-wide uplift and subsidence with peak

vertical movements on the order of 10 m. Geodetic data suggest that the unrest has been driven by changes in

pressure at depths of about 3-4 km or less. The pressure change has been attributed to the intrusion of magma or

to disturbances in geothermal fluids within the shallow crust. A major goal is to determine the relative contribution

of each process, because the potential for eruption is significantly enhanced if magma movement emerges as the

primary component.

Uncertainties in the physical properties of crustal rock at depth are a key source of ambiguity in interpreting

unrest. Of particular interest are better-constrained values of permeability and rock strength, because these

parameters (1) determine the efficiency of fluid flow through geothermal systems and (2) constrain the stresses

required for magma intrusion. Reducing the uncertainties is a primary goal of the Campi Flegrei Deep Drilling

Project (CFDDP), which is supported by the International Continental Scientific Drilling Program (ICDP). A 500-m

deep pilot hole was drilled from July to December 2012, ahead of the main 3.5-km borehole scheduled for the next

year.

An adapted leak off test (LOT) was performed along an 84-m free section at the bottom of the borehole, in

order to measure the in-situ permeability and tensile strength of crustal material (predominantly grey tuff). The

permeability test was performed by injecting water at pressures greater than hydrostatic and then recording the

pressure decrease with time after the water pump had been switched off. The measurements yielded average

permeabilities of 2x10-13 to 4.3x10-13 m2. These values are larger than those obtained from laboratory samples

(about 0.01-0.1 m in dimension) and represent the effective permeabilities at the length scales of at least 1-10

m that are appropriate for modelling the large-scale flow of geothermal fluids. The associated values for bulk

tensile strength were 5.7-6.4 MPa. Such values are typical of the shallow crust and provide upper limits on the

overpressure in a magma body than can be supported by surrounding rock without inducing bulk failure and the

possibility of magma escape.




Contemporaneous eruptions at 4.0 ka from 5.4 km apart vents within Campi Flegreicaldera (Southern Italy): a comparison to Rabaul caldera

Celine Formentraux1, Roberto Isaia2, Mauro Rosi3, Alessandro Sbrana3, Antonella Bertagnini1, PaolaMarianelli3

1Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Pisa, Italy, 2Istituto Nazionale di Geofisica eVulcanologia, Osservatorio Vesuviano, Italy, 3Dipartimento di Scienze della Terra, Universita di Pisa,

Italy


The Campi Flegrei caldera is a collapsed structure mainly resulting from the Campanian Ignimbrite (39ka) and the

Neapolitan Yelow Tuff (15ka) eruptions. Post caldera activity has occurred at vents scattered within the caldera

with eruptions concentrated in relatively short time periods (volcanic epochs) alternating with millennia-lasting

volcanic quiscence. The eruptions at different vents within each epoch have been classically considered to occur

in sequence. However, Isaia et al (2009) documented for the first time that the eruption occurred at Averno and

Solfatara craters, situated 5.4 km apart, actually occurred, at least in part, simultaneously.

In this study we describe and analyses in detail a key tephra section located about 1.5 km NW of the Solfatara

crater where the deposits originated from the two vents are interstratified. The sequence, about 1m thick, consists

of almost plane-parallel alternatng green and pink colored ash beds that overlie a layer of discontinuos pumice

bombs. The green-colored ash beds forms more than half of the succession, the pink-colored ash consists of

several discrete mm- to cm-thick beds. SEM ash clasts characterization coupled with EDS analyses of fresh

glass shards and pumice matrix, indicate that the fine-grained green-colored ash, consists of hydrothermal altered

materials erupted by the Solfatara crater. In contrast, the pink-colored ash consists of fresh glass shards with

composition range similar to the Averno eruption products. In particular the comparison of the glass chemestry

with the one made on the proximal eruptive sequence of the Averno eruption (Formentraux et al., 2012) enabled

us to assess that the studied section mirrors the entire Averno sequence and that the bomb layer at the base was

erupted during the early plinian phase of the eruption.

We conclude that the Averno and Solfatara eruptions started and ended almost contemporaneously. The

comparison with the Tavurvur and Vulcan pyroclastic sequences produced during the very recent eruptive

vents occurred in 1994 at the Rabaul caldera shows striking similarities suggesting that the occurrence of

contemporaneous eruptions at calderas might be more frequent than previously thought. Identifying these type

of event can enhance the volcanic hazards assessment and the expected future eruptive scenarios within active

calderas.

Formentraux C., Metrich N., Bertagnini A., Rosi M. (2012). Contrib. Mineral. Petrol., DOI

10.1007/s00410-012-0720-1.

Isaia R., Marianelli P., Sbrana A., (2009). Geoph. Res. Lett., 36, L21303, DOI 10.1029/2009GL040513.




Is fluid injection the predominant source of Campi Flegrei (Italy) unrests? Clues from thecomparison of inflations and deflations.

Antonella Amoruso, Luca Crescentini, Ilaria Sabbetta

University of Salerno, Department of Physics, Italy


The Campi Flegrei (CF) caldera is a high-risk volcanic area located West of Naples (Italy). CF is generally uplifting

since a few years, after a 20-year-long subsidence started at the end of the recent 1982-84 unrest (maximum uplift

about 180 cm). Few mini-uplifts with subsequent total or partial recoveries are superimposed on the multiannual

trend. While mini-uplifts are usually ascribed to changes in the sub-surface hydrothermal system or injections of

magmatic fluids, a long-standing controversy characterizes the interpretation of the 1982-1984 uplift episode, both

in the source geometry and unrest cause (intrusion of magma, injection of magmatic fluids, or instability of the

hydrothermal system).

As regards the 1982-1984 uplift, deformation data (obtained from leveling and triangulation surveys) seem

consistent with different source models, like pressurized ellipsoidal cavities, mixed-mode faults, fluid injection. As

a consequence, deformation data seem unable to resolve the above-mentioned controversy. However, numerical

models of fluid injection indicate strong time variations in the surficial deformation pattern during inflation and even

more between inflation and deflation.

Here we compare surficial deformation patterns of the major mini-uplifts and subsequent recoveries, the 1995-2000

subsidence, and the 1982-1984 unrest. We use SAR data (ENVISAT ascending and descending orbits, courtesy

of IREA/CNR, Naples, Italy) between 1995-2010, and leveling and triangulation (EDM and angular) data for

the 1982-1984 unrest. Horizontal displacements from triangulations are given with respect to a local not-fixed

coordinate system whose origin coincides with a reference triangulation monument and whose y-axis points toward

a second one. Since the overall deformation pattern is always nearly axial, we have searched for a roto-translation

capable to make the 1982-1984 horizontal displacements radial, thus transforming displacements with respect to

the local coordinate system to absolute displacements.

We find that all the overall deformation patterns, both uplifts and subsidences, coincide within errors and noise

until 2007. The only evident difference between the 1982-1984 unrest and the subsidence phase relates to a small

area (Solfatara) rich of fumaroles. Although we cannot reject fluid injections, the distinct constancy of the surficial

deformation pattern may rule out their predominant role as source of deformation at CF, maybe apart from Solfatara.

After 2007 the deformation pattern seems to consist of the "usual" pattern and an "anomalous" side uplift, whose

origin is under investigation.




Changbai volcano: seismic tomography, origin and East-Asia tectonics

Dapeng Zhao

Tohoku University, Japan


Recent high-resolution seismic tomography revealed a prominent low-velocity anomaly from the surface down

to 410 km depth beneath the Changbai volcano and a broad high-velocity anomaly in the mantle transition zone

under East Asia (Zhao et al., 2004, 2009; Lei and Zhao, 2005). Focal-mechanism solutions of deep earthquakes

indicate that the subducting Pacific slab under the Japan Sea and the East Asia margin is subject to compressive

stress regime (Zhao et al., 2009). These results suggest that the Pacific slab meets strong resistance at the

660-km discontinuity and so it becomes stagnant in the mantle transition zone under NE Asia. The upper mantle

under NE Asia has formed a big mantle wedge (BMW) above the stagnant slab (Zhao et al., 2007, 2011; Zhao

and Liu, 2010). The BMW exhibits low seismic-velocity and high electrical-conductivity, which is hot and also wet

because of the deep dehydration reactions of the stagnant slab and the convective circulation process in the BMW.

These processes lead to the upwelling of hot and wet asthenospheric materials and thinning and fracturing of the

continental lithosphere as well as the formation of the active intraplate volcanoes in NE Asia. Therefore the active

Changbai intraplate volcanism is not related to a deep mantle plume but is caused by the plate tectonic processes

in the upper mantle and the mantle transition zone. A better understanding of these processes can be achieved

by deploying a network of seismic stations on the Korea Peninsula and Japan Sea to determine higher-resolution

mantle tomography under NE Asia in addition to conducting other geophysical and geochemical investigations of

the region.

References

Zhao, D., J. Lei, R. Tang (2004) Origin of the Changbai intraplate volcanism in Northeast China: Evidence

from seismic tomography. Chinese Sci. Bull. 49, 1401-1408.

Lei J., D. Zhao (2005) P-wave tomography and origin of the Changbai intraplate volcano in Northeast Asia.

Tectonophysics 397, 281-295.

Zhao, D., S. Maruyama, S. Omori (2007) Mantle dynamics of western Pacific to East Asia: New insight from

seismic tomography and mineral physics. Gondwana Res. 11, 120-131.

Zhao, D. et al. (2009) Seismic image and origin of the Changbai intraplate volcano in East Asia: Role of big mantle

wedge above the stagnant Pacific slab. Phys. Earth Planet. Inter. 173, 197-206.

Zhao, D., L. Liu (2010) Deep structure and origin of active volcanoes in China. Geoscience Frontiers 1, 31-44.

Zhao, D., S. Yu, E. Ohtani (2011) East Asia: Seismotectonics, magmatism and mantle dynamics. J. Asian Earth

Sci. 40, 689-709.




A preliminary evaluation of the impact of pyroclastic flows using TITAN2D at theBaekdusan volcano according to three different scenarios

Sung-Hyo Yun1, Angelo Paone1, Giovanni Macedonio2, Jeong Hyun Lee1, Sun Kyeong Kim1, YunJeong Kim1

1Department of Earth Science Education, Pusan National University, South Korea, 2OsservatorioVesuviano, Istituto Nazionale di Geofisica e Vulcanologia, Napoli, Italy


The Baekdusan volcano was formed by three stages of activity: (a) a basalt shield(aging between 2.77 and

0.31Ma), (b) a trachytic comendite stratocone(aging between 1.19 and 0.02 Ma), (c) a trachyte-comenditite

ignimbrite deposits(aging between 20 ka up to the present). Volcanic seismicity, ground deformation, and volcanic

gas geochemistry yields new evidence for magmatic unrest of the volcano between 2002 and 2006. The monitoring

data suggest that the Mt. Baekdusan is a potentially active volcano and close attention is needed. One of

the possibly treat of this volcano is the pyroclastic flow volcanic hazard. In order to evaluate the pyroclastic

flow emplacement on this hazardous volcano, we use Titan2d code. It is based on a depth-averaged model for

an incompressible granular material, governed by Coulomb-type friction interactions. The governing equations

are obtained by applying conservation laws to the incompressible continuum, and then taking advantage of the

shallowness of the flows to obtain simpler depth-averaged representations. The motion of the material is considered

to be gravitationally driven and resisted by both internal and bed friction forces. The resulting hyperbolic system

of equations is solved using a finite-volume scheme with a second-order Godunov solver. The DEM file (UTM

easting, UTM northing and elevation in meters) must be properly configured to operate with TITAN2D through

the use of Grass GIS Software. The other input parameters used are: volume(5-10x107 m3, 1x109 m3, 2x1010

m3) of a vertical cylinder pile(maximum height, major and minor axes), center of initial volume(the location of the

initial pile center are the vents of the 1903, 1702, 1668, Millennium eruptions), several orientation angle of the

initial pile were selected, internal friction angle(25-30 degree), bed friction angle(16-25 degree).The pyroclastic

flow run-out calculated in the field are small(3000 m, 1903 eruption) intermediate(5000 m, 1668-1702 eruptions),

large(6-70000 m, Millennium eruption). The initial velocities(m/s) range from 50(1903 eruption) to as high as

300(Millennium eruption). The input parameters have constructed three scenarios(1903, 1702, 1668, Millennium

eruptions) following the recent volcanic history of Baekdusan volcano. These eruptions brace all the possibly

explosive eruptive scenarios that can occur at Baekdusan volcano in the future. Preliminarily, the 1903 type

scenario has been performed, according to the vent location, the flow moves in diverse direction(NE, SE) with

a thickness of 3 m, if the vent is center of the caldera the flow fills the caldera with a thickness of 5 m. The modeling

of the other two scenario are in progress.

Acknowledgments : This research was supported by a grant (NEMA-BAEKDUSAN-2012-1-2) from the Volcanic

Disaster Preparedness Center sponsored by the National Emergency Management Agency of Korea




Eruptive history of Tianchi Volcano, Changbaishan, northeast China: Process and hazard

Haiquan Wei1, James Gill2

1Key laboratory of Active Tectonics and Volcano, Institute of Geology, CEA, China, 2Earth andPlanetary Sciences Department, University of California, Santa Cruz, USA


Tianchi volcano, also known as Changbaishan or Baitoushan volcano, on the China-North Korea border is a

historically active polygenetic central volcano, which consists of three parts: a lower basaltic shield, an upper

trachytic (and comendite) composite cone, and young comendite ignimbrite flows as shown in the table. The

Millennium Eruption occurred between 938 and 946 AD. Three have been some additional, small "historical"

eruptions in the last three centuries.

Between 2002 and 2005, unrest signals including seismicity, deformation, and the helium and hydrogen gas

contents of spring waters increased markedly, causing a regional concern. We attribute this unrest event to magma

recharge or volatile exhalation or both at depth, followed by two episodes of addition of magmatic fluids into the

overlying aquifer without a phreatic eruption. The estimated present magma accumulation rate is rather low to

account for the 2002-05 unrest but there are some indications for magma mixing process happened beneath the

volcano in the past. The most serious volcanic hazards neat Tianchi Volcano are related to ignimbrite-forming

eruptions and lahars.

Stages; Formation and Age; Lithologies

Historical or post-ignimbrite eruptions;

Liuhaojie tuff ring (1903 AD?); Comendite pheatomagmatic layers

Wuhaojie (1702 AD?); White gray comendite fine glass ash

Baguamiao (1668 AD?); Dark gray trachyte ignimbrite and pumice

Late: Ignimbrite-forming;

Millennium eruption (946 939 AD); White gray comendite ignimbrite and air fall pumice and B-Tm ash with minor

trachyte ignimbrite and air fall pumice

Qixiangzhan (17 Ka); Comendite lava and pyroclasts

Tianwenfeng falls (25 Ka); Yellowish fallouts on the rim and B-V tephra layer

Middle:Composite cone construction;

Baitoushan 3 (0.02-0.22 Ma); Trachyte and comendite lava

Baitoushan 2 (0.25-0.44 Ma); Trachyte with Laohudong basalt

Baitoushan 1 (0.53-0.61 Ma); Trachyte

Laofangzixiaoshan (0.75-1.17 Ma) Basalt

Xiaobaishan (1-1.49 Ma); Trachyandesite and trachyte

Early: Shield-forming;

Baishan (1.48-1.66 Ma); Basalt

Toudao (2.35-5.02 Ma); Basalt

Naitoushan (15.6-22.64 Ma); Basanite, Basalt




Major historical eruptions of Tianchi volcano, Changbai Mountain, NE China

Jianli Sui, Qicheng Fan, Ni Li

Institute of Geology, China Earthquake Administration, P.R. China


Tianchi volcano, is a huge stratavolcano located on the border of NE China and North Korea The Millennium

eruption of Tianchi volcnao, erupted at somewhat AD 1000, is one of the world biggest eruptions in the last

2000 years. Historical documents reveal that the last eruption may occurred in about 300 years ago(AD1668

and AD1702). Activity of this volcano is monitored by all modern methods. Potential eruptions of this world most

dangerous volcano attract attentions from both the public and the scientific. However, better understanding of the

historical eruptions of Tianchi volcano will help us to learn more about the future activity. Here we report our recent

studies on major historical eruptions of Tianchi volcano.

Five major eruptions, labeled as T1 to T5, in Holocene have been recognized in Tianchi volcano. The first one

eruption, T1, occurred in about 5000 aBP, formed a thick layer of yellow pumice, widely distributed on and around

the volcanic cone of Tianchi. The second one, T2, usually known as Qixiangzhan eruption, occurred in 2000

aBP, formed a small lava flow of black obsidian, with limited distribution on the northern peaks of Tianchi. T3

is the famous Millennium eruption, the most important eruptions in Tianchi. Volcanic productions, grey pumice,

cover large area from Tianchi volcano to 1000 km eastward of Japan. After the Millennium eruption, several

eruptions have been argued in the past few hundreds of years, and two eruptions, T4(red) and T5 (black) have

been recognized here. T5 may have a small vent at the water vent of Tianchi crater, and also distributed on the top

of Tianchi cone and Bingchang. Mineral assemblage and major element geochemistry of voclanic glass of these

historical eruptions have distinguishable features.

Fractional crystallization in crustal magma chamber dominants the nature of major historical eruptions of Tianchi

volcano. Typical mineral assemblages of fractional crystallization are alkaline feldspar and hedenbergite, with

minor other minerals of fayalite, quartz, apatite and Fe-Ti oxide. E probe data indicate that the historical eruptions

have similar mineral chemistry. A simple model have been used to estimate the extent of fractional crystallization.

If we take a magma of SiO2 64

Supported by NSFC 40972048.




The millennium eruption of Changbaishan volcano in northeast China: High-precisionwiggle-match radiocarbon chronology and implications

Jiandong Xu1, Bo Pan1, Tanzhuo Liu2, Irka Hajdas3, Bo Zhao1, Hongmei Yu1, Ruoxin Liu1

1Institute of Geology, China Earthquake Administration, China, 2Lamont-Doherty Earth Observatory ofColumbia University, USA, 3Laboratory of Ion Beam Physics, ETH, Switzerland


The 10th century AD eruption (or so-called millennium eruption) of Changbaishan volcano, located at the border

between China and North Korea, is one of the biggest (VEI 7) on Earth over the past 2,000 years. However,

the exact timing of the event has been under intense debate for nearly three decades and no unambiguous and

consensus radiometric chronology exists for the eruption at present. In this study, we report a new accurate and

precise chronology of AD 946 ± 3 for the millennium eruption, which was derived from radiocarbon wiggle-match

dating of a 264-year old tree trunk (with bark) buried in the pyroclastic flow deposits of Changbaishan volcano.

High-precision radiocarbon measurements were made on 27 sequentially sampled annual rings of decadal intervals

with analytical precision of ± 25 14C years, on this 260-year tree-ring sequence that covers three consecutive

wiggles around AD 910, AD 785, and AD 730. Since longer dated tree-ring sequence, finer sample resolution

and higher 14C analytical precision all facilitate more and tighter tie-points for better WM dating, our new date is

believed to represent yet the best high-accuracy and high-precision 14C WM chronology for the Millennium eruption.

Moreover, this new age conforms perfectly to the exact date of AD 946 inferred from Korean and Japanese historical

documents and therefore should end the decades-long debate about the timing of the eruption. No stratospherically

loaded sulfate spike that was likely associated with the eruption is found in the global volcanism record from the

GISP2 Greenland ice core, suggesting that the millennium eruption was a Toba-like "ash giant/sulfur dwarf" and

thus had much smaller climatic impacts than implied by its magnitude in the northern hemisphere (compared to the

climatic impact of the 1815 Tambora eruption that also has a magnitude of VEI 7). Our new chronology will serve

as a solid knowledge basis for better understanding of the recurrence interval and eruptive risk of this potentially

most destructive volcano in northeast China.




Petrogenetic history of alkali magmatism at Baitoushan/Changbaishan volcano, China/N.Korea

Frank C Ramos1, James B Gill2, Corey Dimond1, Sean Scott4, John A Wolff3

1New Mexico State University, USA, 2University of California, Santa Cruz, USA, 3Washington StateUniversity, USA, 4University of Wyoming, USA


The Baitoushan/Changbaishan "hot spot" volcano, located along the China/N. Korean border, has erupted a range

of magma compositions from trachyte to rhyolite in the last ~20ky. These magmas retain a variety of distinct mineral

and melt components that result from the interplay of multiple alkali-rich magmas, including alkali basalts, trachytes,

comendites, and a pantellerite. Potential parental alkali basalts, present as satellite cinder cones surrounding the

main edifice, are in U/Th and Th/Ra equilibrium suggesting a lack of initial disequilibrium or old ages. In contrast,

two younger comendite units, found along the northern ridge of the caldera, both have U/Th disequilibria and

one also has Th/Ra disequilibria, attesting to the youthful nature of comendites at this edifice. 87Sr/86Sr ratios

of individual potassium feldspar crystals are uniform in the younger of these comendites but are variable in the

older, which precludes their derivation from the same magma reservoir. In addition, a pantellerite pumice deposit,

located to the east of the main edifice, is intermediate in age between the comendites and has U/Th and Ra/Th

disequilibria. It contains quartz crystals that have Titanium in Quartz temperatures lower than 700 degrees Celsius

and Th/Ra equilibrium probably as a result of remobilization from a cumulate crystal mush. The highly voluminous

comenditic Millenium Eruption (~1000CE), for which Baithoushan/Changbaishan is known, retains extensive U/Th

and Th/Ra disequilibrium in both whole rocks and minerals, including potassium feldspar and quartz, and zircons

and chevkinite crystals with ages of ~10 ka. Minerals have variable Pb isotope ratios suggesting remobilization

from a crystalline mush. By integrating Sr, Nd, and Pb isotope ratios and U-series age constraints of minerals and

magmas, we will address the petrogenetic history of this active magma system. Where possible, we also evaluate

magmatic conditions at which these interactions occurred using Titanium in Quartz geothermometry, thus building

an integrated picture of the time and conditions in which magmatic components are assembled in alkali-rich magma

systems.




Magma system and its eruption processes of the caldera-forming 10th century eruptionof Changbaishan (Baitoushan) Volcano: Inferred from petrological and geochemical

characteristics

Mitsuhiro Nakagawa1, Jumpei Nishimoto1, Tsuyoshi Miyamoto2, Hiromitsu Taniguchi2

1Hokkaido University, Japan, 2Tohoku University, Japan


The 10th century eruption of Baitoushan volcano forming the summit caldera can be divided into three phases,

Phase 1 to 3 in ascending order, with a short dormancy. Although previous studies have considered both Phase 2

and 3 as the 10th century eruption, pumice fall and pyroclastic flow deposits of the Phase 1 are newly recognized

in the limited areas. The age of the Phase 1 might be several tens and hundreds years before the 10th century

eruption, because there exists thin soil layer between Phase 1 and Phase 3. Juvenile ejecta widely ranges from

comendite to shoshonite (SiO2=53-76, Na2O + K2O=6-13). Major ejecta of the Phase 1 and 2 were comendite

pumice, whereas those of Phase 3 were trachyte scoria. Based on the distribution of core compositions of

phenocrystic minerals in a single sample, chemical trends on oxide-oxide diagrams and Nd isotope ratios, it can

be concluded that five magmas were separately present before the eruption, two types of comendite magma with

higher and lower Nd isotope ratio (high-Nd and low-Nd CM magmas), two types of trachyte magma with higher

and lower SiO2 content (high-Si and low-Si TR magmas) and shoshonite magmas. During the eruption of each

phase, magma mingling and mixing of two or three magmas occurred. The 10th eruption began with the withdrawal

from Low-Nd comendite (low-Nd CM) magma (Phase 1), and was followed by eruption from another comendite

magma (high-Nd CM) after a dormancy (Phase 2). During the later of both phases, mixed magma of comendite

magma (low-Nd CM in Phase 1 and high-Nd CM in Phase 2), high-Si TR and shoshonite magmas also erupted.

Thus, it could be speculated that the high-Si TR magma associated with shoshonite magma had injected into

each comendite magma. These two comendite magmas had been exhausted during the two eruption phases.

Thus, low-Si TR magma mainly erupted in Phase 3. The magma was injected by the shoshonite magma to erupt

with small amount of remnant comendite and high-TR magmas. It can be concluded that three eruption phases

of the 10th eruption were related to withdrawal from isolated voluminous three magmas, two types of comendite

and trachyte (low-Si) magmas, respectively. The genetic model of a large silicic magma system must explain the

processes producing several distinct silicic intermediate magmas at the same time as in the case of the 10th

century eruption of Baitoushan volcano.




Tianchi Volcano Millennium Eruption (VEI 7): differentiation processes and timescale

J. Gill1, F. Ramos2, C. Dunlap3

1UCSC, Santa Cruz CA 95064, USA, 2NMSU, Las Cruces NM 88003, USA, 3CRDF Global, ArlingtonVA 22209, USA


Although the first and largest phase of the Millennium Eruption magma was >95% almost aphyric comendite (SiO2

73, Al2O3 >11, FeO*<5), the second phase contains a mafic component that occurs as dark streaks in white

pumice and as isolated mafic scoria. This mafic component extends from trachyte (SiO2 65, MgO=0.1-0.5) through

trachyandesite (SiO2 60, MgO=1.8-2.7) to trachybasalt (SiO2 50, MgO=4.6). Trachyte is the most common. This

mafic component contains diverse glass that is intimately mixed at the mm-scale, and a bimodal mineral population:

Fo50-80 and Fo8-11; En31-43 and En14-20; An14-66 and An4-5. Comendite has more differentiated and more

uniform mineral compositions. Comendite can be derived from trachyte by about 70% fractional crystallization

of Kfs»Cpx>Ilm>Ol and Ap»Chevkinite. The mafic component is slightly more depleted isotopically than the

comendite with 87Sr/86Sr and 143Nd/144Nd at least 0.00005 lower and higher, respectively, and 206Pb/204Pb

about 0.060 lower, based on bulk pumice analyses. Trace element systematics within both comendites and

trachytes demonstrate mixing with something more mafic than either, but relatively little mixing between comendite

and trachyte. The comendite magma was already 10-20 Ka old at eruption based on U-Th disequilibria in its

trace chevkinites and zircons and 226Ra-230Th equilibrium in bulk comendite pumice. However, the mafic mixing

component is younger, with higher 230Th/232Th and greater 226Ra-230Th disequilibrium. Therefore, separate

comendite and trachyte bodies resided beneath the volcano throughout the Holocene, and both experienced a

major recharge event that triggered the Millennium Eruption. The comendite may have received mostly heat from

the recharge event whereas the trachyte mixed thoroughly with the recharging magma.




Changbaishan seismic mornitoring network and recent unrest of the volcano

Guoming Liu

Changbaishan volcano observatory,Jilin province, China


Changbaishan volcano is the largest potential eruptive volcano in China.Over 12 years of continuous monitoring

of Changbaishan volcano by means of volcanic seismicity, ground deformation, and volcanic gas geochemistry,

new evidence for magmatic unrest of the volcano was found during 2002-2005. In this so-called "active period"

the frequency of volcanic earthquakes increased sharply accompanied by earthquake swarms. Active period

was also accompanied by ground inflation, high values of CO2, He, H2, and high ratios of 3He/4He in volcanic

gases released from three hot springs near the caldera rim. The monitoring evidence implies pressurization of

the magma chamber, possibly caused by incremental magma recharge. The ground deformation data from both

GPS and precise leveling are modeled to suggest the corresponding deformation source at 2-6 km beneath the

volcano’s summit, where earthquake swarms were detected in 2002 and 2003. Our findings suggest that the

magma chamber beneath Changbaishan volcano has waked up and resumed its activity after it has remained

dormant since AD 1903.

Changbaishan volcano has remained inactive since 2006, however several abnormal signals have been observed

in recent years that need special attention. In 2009, precise leveling measurement both in the north and west slope

of this volcano show a change of ground deformation mode from inflation to deflation. The water temperature of

Julong hot spring suddenly rose to 77.7°C, about 3°C higher than in 2010, and this abnormal signal persists to

the present. Taken together, all these new phenomena might indicate the beginning of a new "active period". The

episodic unrests probably caused by pulse of mantle magma intruding into the upper crust.

The magma unrest process in Changbaishan volcano from 2002 to 2005 might be considered as a long term

precursor of the potential eruptive activity. However, earthquake swarms and volcanic tremors have not been

observed since 2006, volcanic gas geochemistry also has no obvious abnormal indication, the volcano is still in its

inactive period now.

Key words:volcano; volcanic event;earthquake;precursor




Volcano-tectonic evolution from Cretaceous to Paleogene time in SW Japan

Hiroaki Komuro, Atsushi Kamei

Shimane University, Japan


The eastern margin of the Eurasian continent during the latest Cretaceous to Paleogene is characterized by the

volcanism associated with cauldron swarms and the related plutonism. The SW Japan block was a part of the

Eurasian continental margin in this age. The igneous activity is divided into 4 stages, i.e., stage I (>100Ma), stage II

(100-90Ma), stage III (90-80Ma), stage IV (75-50Ma) and stage V (44-30Ma). The cauldrons in stages I - IV is likely

to make some clusters along the volcanic front in these stages, while the cauldrons in stage V show en echelon

arrangement along the continental margin. This suggests the change of regional stresses during the hiatus of the

igneous activity in the early Paleogene (50-44Ma).

The clustering of cauldrons in stages I - IV may indicate a plume head beneath the cauldron swarm along the

volcanic front. The accretionary prizm along the outer zone of SW Japan shows that the subducted slab lay beneath

SW Japan in that time. A volcanic front ran concordantly with this accretion zone through SW Japan during stage I

- IV, and then migrated northward in stage V.

Each cauldron in stage V exhibits an elliptical or elongated polygonal outline. The long and short axes of a cauldron

represent the directions of horizontal maximum and minimum compressive regional stresses, respectively. The

elongated shape of a collapse caldera, which is topographic expression of a cauldron, presumably reflects regional

stresses in the lithosphere as well as the geometry of the magma body.

The cauldrons in stage V show a set of right-handed en echelon arrays oriented at an angle of 18 - 20°to the

direction of the volcanic front in this stage. The long axes of some cauldrons, representing the directions of the

horizontal maximum compressive stresses, trend to 20°or more anticlockwise to the cauldron array. This orientation

indicates that SW Japan in stage V was under left-handed Reidel shearing.

Tectonic regime during stage I - IV was probably characterized by the normal subduction followed by plume

ascending along the subducting slab, whereas that in stage V was identified as the oblique subduction yielding

the left-handed lateral movement in the overlain lithosphere.



3W_3C-P1 Date/Time: July 23 Poster

Spatiotemporal variations of geochemical characteristics of volcanic rocks from Asovolcano, SW Japan

Taro Shinmura1, Yoji Arakawa2, Isoji Miyagi3, Masaya Miyoshi4

1Kumamoto Gakuen University, Japan, 2University of Tsukuba, Japan, 3Geological Survey of Japan,AIST, Japan, 4University of Fukui, Japan


Chemical compositions and isotopic ratios (87/86Sr and 143/144 Nd) for late-Pliocene to Quaternary volcanic

rocks in and around Aso area in SW Japan, were determined to reveal spatiotemporal variations of the magma

characteristics and a mechanism of magmatic system including large caldera volcano related to super colossal

eruption.

In this study, volcanic activities in Aso area are divided to 5 stages depending on ages and characteristics of forms

of eruptions. 1) Early pre-caldera stage: HMA (High Mg andesite) (3.9Ma) and High Mg picritic basalt (2.9Ma) are

distributed on southwestern somma of Aso caldera. 2) Late pre-caldera stage: Pre-Aso volcanic rocks (Watanabe et

al., 1989) exposed on caldera wall and somma (0.9-0.4Ma). 3) Caldera forming stage 1st to 4th: 4 colossal to super

colossal eruptions (ca. 270, 140, 120 and 90 ka (Watanabe, 2001)) with voluminous pyroclastic flows including

scoria and pumice. 4) Inter-caldera stage: Volcanic activities between colossal to super-colossal eruptions, mainly

consisted of andesite lava flows (distributed on somma and outside of caldera). 5) Post-caldera stage: Volcanic

activities after Aso-4th (the final caldera forming eruption) within caldera, forming Aso central volcanic cones with

basaltic to rhyolitic volcanic products. Although Nekodake volcano is one of the Aso central volcanic cones and

some volcanic products were dated as in post-caldera stage, geochemical characteristics are strongly related to

volcanic rocks of late pre-caldera stage and basement rocks (granodiorite).

Volcanic products of both inter-caldera and caldera forming stage were characterized by high K2O contents and

same REE patterns and isotopic ratios. On the 87/86Sr-143/144 Nd diagram, isotopic data of volcanic rocks in early

pre-caldera stage separately distributed in two areas (component A: around (0.7040, 0.51285); component B:

around (0.7044, 0.51270) and of both late pre-caldera and post-caldera stage distributed between component A

and B along mantle array. Sr isotopic ratios of volcanic products of both caldera forming and inter-caldera stages

were distributed in very narrow range (0.7040 - 0.7042). Source materials of volcanic products of each stage are

inferred from these results.

A part of this research project has been conducted as the regulatory supporting research funded by the Secretariat

of Nuclear Regulation Authority (Secretariat of NRA), Japan.




Lateral magma intrusion from a caldera-forming magma chamber: Constraints fromgeochronology and geochemistry of volcanic products from lateral cones around the

Aso caldera, SW Japan

Masaya Miyoshi1, Taro Shinmura2, Hirochika Sumino3, Yasuo Miyabuchi4, Toshiaki Hasenaka5, YojiArakawa6

1Geol. Lab., Univ. Fukui, Japan, 2Fac. Econ., Kumamoto Gakuen Univ., Japan, 3GCRC., Univ. Tokyo,Japan, 4Fac. Edu., Kumamoto Univ., Japan, 5Grad. Sch. Sci. Tech., Kumamoto Univ., Japan, 6Grad.

Sch. Life Environ. Sci., Univ. Tsukuba, Japan


We investigated the K-Ar ages, and the petrological and geochemical features of lava units from lateral cones

(Omine and Akai volcanoes) and lava distributed area around the Aso caldera in central Kyushu, Japan, in

order to constrain the spatial range of lateral magma intrusion during the caldera-forming stage. The results

of K-Ar age determination showed that most of the analyzed lava units erupted almost simultaneously with the

Aso caldera-forming pyroclastic eruptions (266 to 89 ka; Matsumoto et al., 1991). In addition, the petrography,

major- and trace element compositions, and Sr isotope ratios of these lava units are indistinguishable from the

caldera-forming pyroclastic products. In particular, the decrease in Sr isotope ratios over time observed in these

lava units is consistent with that of the caldera-forming pyroclastic products. The contemporaneous activities of

compositionally similar magmas inside and outside of the caldera presumably indicate the occurrence of a lateral

intrusion of caldera-forming magma, which had accumulated in a huge magma chamber beneath the caldera

system. In the Aso volcano, it is thought that a total of 6.3 volume percent of caldera-forming magma migrated

more than 20 km from the center of the caldera.




Unusually high-temperature andesitic magma erupted shortly before the Aso-2pyroclastic flow from Aso caldera, Japan

Koshi Nishimura1, Yasuo Miyabuchi2, Hirohito Inakura3, Tetsuo Kobayashi4

1Natural Science Laboratory, Toyo University, Japan, 2Faculty of Education, Kumamoto University,Japan, 3West Japan Engineering Consultants, Inc., Japan, 4Graduate School of Science and

Engineering, Kagoshima University, Japan


Aso Volcano, which is located in central Kyushu, southwestern Japan, is one of the largest caldera volcanoes

in the world. The caldera (25 km north-south and 18 km east-west) was formed by four large pyroclastic flow

eruptions: Aso-1 (270 ka), Aso-2 (14 ka), Aso-3 (12 ka) and Aso-4 (89 ka). Each pyroclastic flow deposit is divided

into several flow units. These flow units were deposited by pyroclastic eruptions which accompanied tephra fall just

before and/or after the generation of large pyroclastic flows, except for Aso-4 eruption. In particular, the stratigraphy

of Aso-2 eruption deposit is very complicated and the chemical compositions remarkably vary between subunits in

the eruption cycle. Andesitic lava flows were generated shortly before the main Aso-2 eruption. Tamaraigawa lava

distributed east of Aso caldera and Iwato, Akita and Togawa lavas distributed west. These lavas are nearly aphyric

and display textural characteristics similar to pahoehoe lava flows.

Lava compositions, temperatures and viscosities have been investigated. All lavas are phenocryst-poor, consist

of two pyroxene andesite and have 61 wt.% SiO2 except for Togawa lava (58 wt.% SiO2; Matsumoto, 1974). The

magma temperatures were estimated using the two-pyroxene thermometer of Anderson et al. (1983). Five to seven

coexisting pairs of pyroxenes were analyzed for each lava. The calculated magma temperatures are 1123±23 °C

(Tamaraigawa lava), 1081±17 °C (Iwato lava), 1061±18 °C (Akita lava) and 1045±24 °C (Togawa lavas). We

calculated the melt viscosity of the Tamaraigawa lava at 1123 °C (two-pyroxene temperature) using the model of

Giordano et al. (2008). The results show that the Tamaraigawa lava has a viscosity lower than 104.5 Pa s (dry

case). Dissolution with water further decreases the melt viscosity to 102.7 Pa s at 2 wt.% H2O.

In general, andesitic magmas have eruption temperatures of 900-1000 °C and viscosities around 108-109 Pa s.

The andesitic lavas in this study, therefore, had unusually high temperature and low viscosity conditions similar to

basaltic lavas. The data are consistent with the textural characteristics of pahoehoe lava flow.




Volcanic history of western margin of Aira caldera based on K-Ar geochronology

Masafumi Sudo1, Kozo Uto2, Daisuke Miki3, Kazuhiro Ishihara3, Yoshiyuki Tatsumi4

1University of Potsdam, Institute of Earth and Environmental Science, Germany, 2National Institute ofAdvanced Industrial Science and Technology, Japan, 3Kyoto University, Disaster Prevention Research

Institute, Japan, 4Kobe University, Graduate School of Science, Japan


Western margin of Aira caldera consists of the Yoshino-dai plateau with a gentle slope to the west and the northern

high volcanic mountain area with up to 550 m height with a steep caldera wall faced to Kagoshima Bay of more than

200 m height. It is the area where the most voluminous volcanic rocks distribute around the rim of Aira caldera as

well as the post-caldera Sakurajima volcano. Therefore, it is an important area to understand how the Aira caldera

has been evolved.

Systematic K-Ar dating by sensitivity method at Kyoto University and Geological Survey of Japan has been

performed to the volcanic rocks distributed in this western margin of Aira caldera. The history of the pre-caldera

volcanic activities obtained is as follows: (1) Andesitic magmas erupted at the present high volcanic mountain

area in the north of present Yoshino-dai plateau from 1 to 0.7 Ma, then formed the volcanic bodies with up to

550 meters height. (2) Next, dacite erupted at the southeastern side of the above mountain area at 0.5 Ma,

besides, subsequently basaltic lavas erupted from the present Kagoshima Bay side, overlied the dacites and had

also flown between volcanoes of the high volcanic mountain area and in the Kagoshima Bay. (3) Then, rhyolitic

pyroclastic flows and subsequent basaltic lava flows erupted westward from the Kagoshima Bay side again and

formed Yoshino-dai plateau from 0.45 to 0.35 Ma. (4) The volcanic body which located in the Kagoshima Bay

and erupted magmas westwards through 0.5 to 0.35 Ma had collapsed and disappeared some time after 0.35 Ma.

Then the Yoshino-dai plateau and the caldera wall had remained as the foot of the previous volcano as shown in the

present time. It was also found that the basaltic magma erupted around Aoshiki which locates in the 10 km north

from the Yoshin-dai plateau at 0.08 Ma. This implies that the activity of basaltic magma occurred during the period

when the volcanic activity at Aira caldera became more active after 0.1 Ma than before as shown in the multiple

pyroclastic eruptions including Aira pyroclastic eruption. The K-Ar ages obtained so far from the whole Aira caldera

area implies that the location of volcanic activity had migrated along the caldera rim since 1.5 Ma.




Holocene uplift of Aira caldera, southern Japan

Hiroshi Moriwaki1, Mitsuru Okuno2, Toshiro Nagasako1, Akio Ohira3, Yoshiaki Matsushima4

1Kagoshima University, Japan, 2Fukuoka University, Japan, 3University of Miyazaki, Japan, 4KanagawaPrefectural Museum of Natural History, Japan


Aira caldera, accompanied by Sakurajima volcano on its southern rim is filled with sea water, forming the inner part

of Kagoshima Bay. Emerged sea-level records, i.e. emerged marine deposits and marine terraces occurring around

Aira caldera, provide excellent data for obtaining the Holocene crustal movement of the caldera. We obtained the

mode of crustal movement of Aira caldera in the past c. 7,000 years on the basis of the elevations of those marine

records and their chronology using C-14 age and tephra beds. The highest sea-level at c. 7,000 years cal BP

recognized in the northern to northwestern rim of the caldera, attains more than 10 m above the present sea level.

The distribution of the sea-level records clearly indicates upwarping with a center of slightly western part of the

caldera. The mean uplift rate, c. 1.4mm/yr and the mode of upwarping nearly coincide with those of the historical

uplift associated with the volcanic activities of Sakurajima volcano during the past c. 500 years, suggesting that

the Holocene upwarping of Aira caldera reflect the volcanic activities of Aira caldera including Sakurajima volcano.

This means that the findings of Holocene crustal movements recorded in the coastal deposits and landforms are

important for evaluating the future volcanic activities of Aira caldera and Sakurajima volcano.




Cycles of magma activities leading to catastrophic eruptions in aira caldera in Kyushu,Japan

Yuko Sekiguchi1, Toshiaki Hasenaka2, Shinji Nagaoka3, Yasushi Mori4

1Sapporo District Meteorological Observatory, Japan, 2Grad School of Science and Technology,Kumamoto University, Japan, 3Faculty of Education, Nagasaki University, Japan, 4Kitakyushu Museum

of Natural History and Human History, Japan


Aira pyroclastic eruption (29 ka; 450 km3), one of the largest caldera-forming events in Japan, was proceeded

by continuous tephra eruptions (100–30 ka), and provides a unique opportunity to examine transition of magma

reservoir compositions leading to the caldera-forming eruption. We report magma compositions between 100 ka

and the present, and propose cycles of magma activities in which the mafic magma activity marks early stage

and the felsic one marks late stage, ending with catastrophic eruption. We classified one mafic magma group and

two felsic magma groups on the basis of mineral assemblages, group-M (under 59 wt.% SiO2; <95 ka) containing

plagioclase, two pyroxenes and rare olivine, group-F1 (63–70 wt.% SiO2; 95–85 ka) containing plagioclase, two

pyroxenes and hornblende and group-F2 (73–78wt.% SiO2; 60–30 ka) containing plagioclase, orthopyroxene and

quartz. Products of Aira pyroclastic eruption, belong to Group-F2 magma.

Patterns of transition of magma compositions during 100 ky at Aira revealed that Group-M mafic magmas

were active before felsic magmas (F1 and F2). Aira pyroclastic eruption marks the final eruptive event of Group-F2

activity. Fukuyama pumice fall eruption, which is the largest eruptive event (about 40 km3) before Aira event (29

ka), marks the final eruptive event of group-F1 magma activity. Incompatible trace element compositions show that

Group-M magma and Group-F2 magma do not represent parent-daughter relationship. Contents of incompatible

elements of Group-F2 magma increase with time (SiO2: 2–4 wt.%; Rb: 5–20 ppm) from 60 ka to 30 ka, to the

composition similar to that of Aira pyroclastic eruption. Compositional variations observed among Group-F2

magma are explained by crystal fractionation of the mineral phases contained in the parent magma. Felsic magma

similar in composition to Aira pyroclastic products appeared 1,000 years before the event.

Volcanic products from Sakurajima volcano (25.5 ka–the present), show binary magma mixing between

basalt and dacite. Their mafic end member is compositionally similar to the Group-M magma which appeared in

the first and the second cycles. Neither F1 nor F2 magmas are possible candidates of felsic end member of mixing.

It is implied that different felsic end member magma, i.e. F3, exists in magma reservoir beneath the present Aira

caldera. Magma activities of Sakurajima volcano, probably forms the felsic stage of the new cycle.




Caldera structure of Izu Oshima Volcano, Japan. revealed by new drilling survey

Yoshihisa Kawanabe

Geological Survey of Japan, AIST, Japan


Izu-Oshima volcano is basaltic stratovolcano with a summit caldera. The caldera is high gravity anomaly type and

the eastern rim is almost buried by the younger deposits. It is thought that this caldera was formed by collapse with

steam explosion about 1700 years ago. Some drilling survey, such as Isshiki et.al.(1963) were done in western

part of the caldera but there are no survey in eastern part. To investigate the structure of the basaltic caldera, we

conducted a new drilling survey of 100m depths in the eastern part of caldera.

The proportion of lava in the core is 42% and pryroclastics is 58%. 7 lava units are found in the lower and the upper

part of the core except the 4th unit. Thick volcanic breccia and ash fall deposit with accretionary lapilli are found

in 38m to 50m deep. This lithology and stratigraphy is very similar to the products at the time of the latest period

of caldera formation. The trace element composition of the lava also have different element ratio above and below

this breccia, same as the surface distributed lava, thus it indicates this layer is the latest caldera forming deposits.

So, the caldera floor hight is about 400m above sea level, 60m shallower than expected from Isshiki et.al.(1963).

From lithology and 14C dating, the eastern caldera had been in the environment that pyroclastic fall only deposited

and lava did not come about 5,000 years ago until several hundred years ago. From these, at least three caldera is

assumed in Izu-Oshima volcano, the younger western caldera formed about 1700 years ago and the older western

caldera about 5,000 years ago and the oldest eastern caldera. Using the estimated spread of the younger western

caldera by this study and the caldera floor altitude by previous research, the volume of caldera filled material is

estimated. The estimated volume is about 1.6 km3 and using the lava-pyroclastic ratio (75:25; Isshiki et.al. 1963),

the volume of lava is estimated about 1.2 km3.




The origin and magmatic evolution of Quaternary lavas of Sakurajima volcano, southernKyushu Island, Japan: inferred from Geochemical and Sr-Nd-Pb isotopic constraints

Tomoyuki Shibata1, Jun Suzuki1, Masako Yoshikawa1, Tetsuo Kobayashi2, Daisuke Miki3, KeijiTakemura1

1IGS, Kyoto University, Japan, 2Graduate School Sci. Eng., Kagoshima University, Japan, 3SVRC,DPRC, Kyoto University, Japan


Sakurajima volcano represents a post-caldera volcano linked with the Aira caldera, situating at the volcanic front

of the Ryukyu arc, southern end of Kyushu Island, Japan (Fukuyama, 1978), where the Philippine Sea Plate

(PSP) is subducting, and is considered to be one of the most suitable volcanoes for studying the links between

caldera formation and catastrophic silicic magmatic activity. We collected 25 lava samples, from almost all of the

volcanostratigraphic units (Fukuyama and Ono, 1981), for a major and trace element geochemical and Sr-Nd-Pb

isotopic petrogenetic investigation of Quaternary lavas of Sakurajima volcano.

The Sakurajima lavas are porphyritic andesites or dacites contains Opx-Cpx-Pl with or without Ol as phenocrysts.

The Nb depletion along with enrichments in Rb, K, and Pb show the typical island arc magma characters causing

by the addition of aqueous fluids to the mantle wedge. The Sr, Nd, and Pb isotopic compositions plot close to a

mixing curve between MORB-type mantle and sediments of the Philippine Sea Plate, (PSP) but displaced a bit

towards more radiogenic compositions. Plots of Zr v.s. Nb concentration yield a linear trend that falls on a mixing

line between the values for MORB and average continental crust. These observations indicate that the primary

source magmas were initially generated by partial melting of MORB-type mantle hydrated by fluids derived from

the subducting PSP. The contribution of crustal material during magma evolution is also evident from the Zr/Nb

ratios and Sr-Nd-Pb isotopic compositions. The mixing relations of Sr-Nd-Pb isotopic compositions suggest that

the sedimentary rocks of Shimanto Group can be a source of the crustal materials. Although most of the major

element compositions show a single linear trend on each of the Harker diagrams, two different trends are discernible

on each of the P2O5, and TiO2 v.s. silica diagrams, and are subdivided into low-P and high-P geochemical groups.

The magma mixing trends of Sakurajima lavas, which seem to be extended from mono andesitic end-member to

two different deictic end-members, are observed from the relationships of major element contents and 87Sr/86SrSr

ratios. In addition, the low-P versus high-P groups of lavas show distinctive distribution patterns, whereby the high-P

lavas are surrounded by low-P lavas in the central to southern parts of the Sakurajima volcano. These observations

indicate that mixing of andesitic and dacitic magmas played an important role in the genesis of lavas of Sakurajima

volcano, and that multiple dacitic magma chambers with different geochemical characteristics once existed beneath

the Sakurajima area at relatively shallow levels in the crust. From the relations between SiO2 and Sr isotope ratios,

an AFC process is required to originate the andesite and dacite end-members.




Near-vent morphology and dispersion timing of the climactic PDC in 7300 BP marinecaldera formation of Kikai caldera in southern-off Kyushu Island, through seismic

reflection survey

Fumihiko Ikegami1, Shoichi Kiyokawa1, Hisashi Oiwane2, Yasuyuki Nakamura3, Katsura Kameo4, YutoMinowa1, Takashi Kuratomi1

1Department of Earth and Planetary Sciences, Kyushu University, Japan, 2National Insitute for PolarResearch, Japan, 3Japan Agency for Marine Earth Sciences and Technology, Japan, 4Atmosphere and

Ocean Research Institute, Tokyo University, Japan


Kikai caldera (Matsumoto, 1943) is a mostly submerged highly active caldera complex located in the southern

Japan 40 km off Kyushu Island. The formation of the Kikai is believed to be responsible for dispersion of Akahoya

tephra (Machida and Arai, 1978) in 7300 cal. BP (Fukusawa, 1995). The eruption is characterized by intense

earthquakes (Naruo and Kobayashi, 2002), low-aspect ratio Koya-Takeshima PDC (pyroclastic density current;

Maeno and Taniguchi, 2007) and tsunami (Geshi, 2009) which were taken place at the climax of the eruption.

We conducted seismic reflection observations in two survey cruises (KT-10-18 and KT-11-11) in 2010 and 2011

using a research vessel Tansei-maru of JAMSTEC (Japan Agency for Marine-Earth Science and Technology). The

sound source was a 105+45 cubic inches G-I gun with 10 seconds of shot interval, and a 48-channled 1.2 km-length

streamer cable was used for acquisition. Totally 24 profiles were obtained with the speed of 4 knots.

1. Possible climactic PDC deposits

Facies in Kikai area show clear distinction between caldera outskirts, caldera basin, and central rise of the caldera.

At the caldera outskirts, there is a thick (around 100 m) acoustically chaotic layer named A3 in our interpretation that

alike to other presumed PDC-originated deposits (e.g. Lebas et al., 2011). The layer is covered by some intensively

stratified deposits except on the southern steep slopes. In contrast with outskirts however, equivalent considered

layer in the caldera basin shows acoustic transparency with slight stratification within it, and the bottom terrain of it

marks major unconformity. Neither of such characteristic layers was observed at the center of the caldera where

large topographic rise exists.

2. Intrusive structure along the caldera-rim

Volcanic bodies are distributed along the caldera-rim. The largest one at the southeastern end has 2.5 km wide,

and the rolling-up horizons beneath of it indicate it occurred just before the climactic dispersion. Presence of such

bodies make difficult to evaluate caldera displacement however, it reaches 400 m in maximum at the rim of east to

southeast.

3. The timing of the climactic dispersion and caldera collapses

Both caldera displacements and inner fractures seem to cut every deposit at every direction, therefore the caldera

collapse should be occurred enough after the climax of the eruption. Enigmatic shortage of the talus deposits may

reflect the subsidence and opening of the fractures happened gradually.




Frequency of a caldera forming eruption occurred in the Kyushu Island on the basis ofthe high-resolution tephra stratigraphic record in the Kinki district, central Japan

Yoshitaka Nagahashi

Fukushima University, Japan


Extremely a large-scale explosive eruption, for example a caldera forming eruption, it is sometimes described

that such eruption occurred once every 10,000 years. We have been constructed the high time resolution tephra

stratigraphic record for Pliocene to Holocene strata in Japan.

A framework for the Pliocene to Holocene tephra stratigraphy in the Kinki district has been established by

litho-stratigraphy of the Plio-Pleistocene Osaka Group in and around the Osaka Plain, in addition, by using a

number of drilling cores sediments in the Osaka Group and from Lake Biwa. The Osaka Group is mainly composed

of fluvial sand, gravel and silt intercalated with twenty-one marine clay beds and over eighty tephra beds. It is

especially significant that marine clay beds had been deposited by the corresponding sea level change of the

glacial / interglacial cycles. Depositional age of the Plio-Pleistocene tephra beds is determined by a combination

of lito-, bio- and magneto-stratigraphy and radiometric age of tephra beds. In particular, staratigraphic positions of

tephra beds in the past one thousand three hundreds years is accurate as the stratigraphic sequence because of

corresponding the twenty-one marine clay beds to the Oxygen Isotope Stage 37 to Stage 1.

Most of the tephra beds can be individually traced to their source volcanoes based on the volcanic glasses

chemistry using the microprobe (EDS) analysis and petrographic properties. As a result, we are identified nineteen

co-ignimbrite ash beds during the last 130,000 years. these tephra beds have been mainly derived from Quaternary

caldera volcanoes in the Kyushu district, named as the Shishimuta, Aso, Kakuto, Aira, Ata, and Kikai calderas. The

excellent tephra stratigraphic record also shows the relation of the stratigraphic positions to the Oxygen Isotope

Stage. In short, a many of the co-ignimbrite ash beds have emplacement at the low sea-level stand, and it

is possible that the timing of eruption is related to local tectonics and hydro-isostatic stress in the upper crust

corresponding to the sea level changes.




Modern iron sedimentation and hydrothermal activity at post Kikai Caldera volcano inSatsuma Iwo-Jima, Kagoshima, Japan: To understand modern bedded iron formation at

shallow hydrothermal environment

Shoichi Kiyokawa, Takuya Ueshiba, Yuto Minowa, Tomoki Nagata, Tomomi Ninomiya

Kyushu university Earth and Planetary Science, Japan


Satsuma Iwo-Jima Island at northwest limb of the Kikai Caldera was located 38 km south of Kyushu Island, Japan.

Nagahama Bay in the island took on reddish brown color and high density iron-oxyhydroxides seawater which is

formed by the neutralization of ferrous iron in acidic hot-spring. After dredging work at 1998, iron-oxyhydroxide

sediments have accumulated about 1 to 1.5 meter on the bottom of this bay. For estimate the amount of iron

material discharge and sedimentation history, we collected 13 core samples from this bay and form cross section in

this bay. The stratigraphy in these cores well preserved volcanic activity and weather condition such as storm, heavy

rain and calm condition from comparison meteorological data. Also cross section and column show elucidated

sedimentation period of characteristic key beds and calculated sedimentation rate of iron-oxyhydroxides.

Cores contains iron-oxyhydroxides muds layers, three thick tuff beds and a thick sandy mud bed. Iron-oxyhydroxides

mud consisted minor volcanic glass and mainly 1 micro meter to 100 nano mater Fe mineral. Tuff beds were

composed of volcanic glass and Si-bearing minerals. Sandy mud bed was essentially a mixture of rock fragments,

volcanic glass and fine reddish-brown grains. Comparison with this stratigraphy and meteorological data show that

ash beds were correlated to heavy rainfall in 2000, 2001 and 2002 which years occur volcanic ash fall at mountain,

and a thick sandy mud bed was corresponded to strong typhoon events in 2004 to 2005. Heavy rainfall supplies

ash material to bottom of seafloor from rhyolite volcano Iwo-Dake depositing unformed tuff-rich sediment which was

deposited at the top of volcano during. Strong typhoon drives Al and Si-bearing material to Nagahama Bay and

these materials are re-sediment together as sandy mud bed. We will try to identify iron formation mechanism to

identified relationship between hydrothermal activity and sedimentation. Before estimation of iron formation, at least

we subtract wave and rain effects. Unformed iron-oxyhydroxides mud accumulated rapid speed at 33.3 cm par year

at 1m-sediment-trap cores placed on the seafloor. From meteorological comparison dating, estimated accumulation

rate of iron-oxyhydroxides mud is 4.7 cm par year. As a result, we estimate that approximate 800 cubic centimeter

iron-oxyhydroxides discharge from bottom of seafloor, and only 14 percent preserved in this Nagahama Bay.




The stratigraphy of Pliocene to middle Pleistocene pyroclastic deposits, MiyagiPrefecture, Japan

Takahiro Kuzumaki1, Akihiko Fujinawa2, Tsukasa Ohba3

1Graduate School of Science and Engineering, Ibaraki University, Japan, 2Department of Earth andEnvironmental Sciences, Ibaraki University, Japan, 3Faculty of Engineering and Resource Science,

Akita University, Japan


In the Kurikoma geothermal area on the frontal zone of the Northeast Japan arc, a caldera cluster has been

formed during recent 9Ma, through the repeated explosive volcanisms which numerous tephra layers in the east of

the cluster. Among the tephra, the latest 4 pyroclastic flow layer; Ikeduki, Shimoyamazato, Nizaka, Yanagisawa tuff

layers have been well determined their localities of vents along with their eruption ages ranging from 0.3 to 0.04Ma.

However, the tephra layers preceding to the above 4layers have not been well defined in terms of their

lithostratigraphy, eruption ages, or localities of eruption vents. These have been regarded as the members of the

Pliocene to middle Pleistocene Onoda Formation. We reported the stratigraphy of the eruptive product in the Onoda

Formation: Takatama (3.3±0.3Ma), Shimatai, Yubama, Chijimisawa (1.00±0.06Ma), Mozume (1.08±0.13Ma and

0.62±0.10Ma), Toshojisawa (0.87±0.21Ma), and Uguisuzawa pyroclastic flow deposits in stratigraphic order with

Aonosawa tephra layers at the top (Kuzumaki and Ohba, 2010, 2011).

In this study, the stratigraphy of pyroclastic deposits in Onoda Formation are reexamined on basis of newly identified

deposits, and the frequency of pyroclastic eruption are examined.

At the 4 newly discovered localities (St, Ot, Tt, and Kt) 7, 5, 3, 2, distinct tephra layers are identified. The layers Kt-1

and -2 and Tt-1, -2, and, -3, in are to be stratigraphically lower than the Yubama pyroclastic flow deposit, because

Kt-1 and -2 are overlain by Yubama pyroclastic flow deposit, and the topography of deposition for Tt-1, -2, and,

-3 is lower than Kt-2. Since St-7 is correlated petrologically with Mozume pyroclastic flow deposit, six pyroclastic

deposits, St-1~-6, are stratigraphically higher than Mozume pyroclastic flow deposit.

At least seventeen distinct pyroclastic deposits above the Chijimisawa pyroclastic flow deposit of ca. 1.0Ma are

identifiable in this area. Therefore, the frequency of pyroclastic eruption is higher than 1 in every 0.06Ma.




Volcanic history of the Takahara Volcano, Northeast Japan arc: Inference of the calderaactivity

Kensuke Tsurumaki

Graduate school of Geography, Meiji University, Japan


Introduction

Caldera activity has large effect on environmental and geomorphic change. Therefore, clarification of the volcanic

history and evolution of magmatic processes is necessary for a long-term prediction.

In Southern part of the Northeast Japan arc, which is south than Fukushima, most of the caldera activity is before

Calabrian, and they are not remarkable expect Takahara volcano after Middle Pleistocene.

Takahara volcano currently formed in northern Kanto plain about 140 kilometers away from the Tokyo, and the

Shiobara caldera about 7 kilometers in diameter existence in northern foot of Takahara volcano, which formed at

0.3Ma in former research. Then, large andesitic stratovolcano with an altitude of nearly 2,000 meters formed after

0.2-0.3Ma. A part of caldera is filled with the lava from stratovolcano. The central part of stratovolcano eroded very

much, it is suggest that there is no eruption from summit recently. By contrast, a dacitic central cone formed in the

Shiobara caldera at Holocene and fumarolic activities around central corn are still continuing. Thus, the two types

volcanism recognized in Takahara volcano. However, the volcanic history is not clarification because there is only

insufficient informational about stratigraphy and radiometric age.

The purpose of this study, therefore, to clarify volcanic history and evolution of magmatic system of Takahara

volcano based on the whole rock chemical analysis and radiometric age determination of Takahara volcanic rocks.

Results

In the pre-caldera stage (<0.6Ma) of Takahara volcano, tholeiitc basalt to andesite volcano was mainly formed with

multiple magmatic systems.

In the caldera forming stage (0.6-0.3Ma), The marker tephra and radiometric age of ignimbrites from Takahara

volcano indicated that several ignimbrites erupted in 0.6Ma and 0.3Ma which is a formation factor of Shiobara

caldera. When the caldera formed for the first time about 0.6Ma, the two ignimbrites erupted from multiple

magma chambers has different chemical composition, which are called Kanawazaki PFD and Katamata PFD.

These ignimbrites widely covered the Nasunogahara region in northern Kanto plain.

After that, basalt to rhyolite lava flows erupted from tholeiitic magma chamber. At the second time of the caldera

formation, Tanohara PFD erupted about 0.3Ma. Then, the magmatic system extremely changed at post caldera

stage (>0.3Ma), and andesitic to dacitic stratovolcano built by newly supplied calc-alkaline magma. The eruption

style of stratovolcano is non-explosive because it is composed mainly of lava flows and intercalated with small-scale

scoria beds.

The center corn erupted at the Holocene is characterized by biotite. It was not contained in volcanic rocks of

post-caldera stage indicates that the high water content magma recently is supplied.

These results infer that the difference in eruption styles was caused by complex magmatic systems in Takahara

volcano.




Post and pre caldera eruption history of Miyakejima Volcano - Entombment process ofHachodaira Caldera and eruption history of pre A.D. 2000 Caldera -

Teruki Oikawa, Nobuo Geshi

Geological Survey of Japan, AIST, Japan


Collapse caldera is a common volcanic structure in basaltic volcanoes. However, the entombment process of

the caldera and the eruption history of the post-caldera are poorly understood. To make an effective forecasting

of the volcanic activity, we have to understanding the variations of the eruption activities during the caldera filling

stage based on the eruption history of the post-caldera period of many volcanoes. Based on the caldera wall

observation, tephrochronology, and radiocarbon dating, in the Miyakejima Volcano, we clarify the entombment

process of Hachodaira Caldera and the eruption history of the pre A.D. 2000 Caldera.

The cross section of Hatchodaira Caldera, forming age at ca. 2.5 ka, is exposed on the wall of the A.D. 2000 caldera

on Miyakejima volcano. The cross section of Hatchodaira Caldera is divided into 7 units; pyroclastic cone (130 m

thick), many thick lava flows (200 m thick), single lava flow (5-10 m thick), scoria fall deposit (40 m thick), lava flows

(30 m thick) and scoria fall deposit, in ascending order. The caldera was filled mainly by the lava flow which makes

a lava lake. On the other hand, many large-scale fissure eruptions, such as Kazahaya Scoria at 1.4 ka (ca. the 6th

century, radiocarbon ages calibrated to calendar years) occurred outside of caldera during the caldera-fill stage (ca.

2.5 ka to the 9th century) of Hatchodaira Caldera. The production rate during the caldera-fill stage of Hatchodaira

Caldera is estimated at least 0.4 km3 / ky. This value is larger than the production rate 0.2 km3 / ky for the last 1100

years after the caldera filling stage.

After the 9th century, the eruption number of pre-A.D.2000 Caldera is 17 times. These eruptions are flank fissure

eruption in most cases. The frequency of eruption is 1-2 times in 100 years. Just before the formation of the

A.D. 2000 Caldera, the frequency of eruption was slightly larger (3 times in 100years). But it is not a significant

difference.




Were the Pre Lower Pumice 1, Lower Pumice 1 and Lower Pumice 2 eruption sequencessourced from the same magma reservoir, Santorini Caldera, Greece?

Jack M. Simmons1, Ray A.F. Cas1, Timothy H. Druitt2, Christopher Folkes1

1School of Geosciences, Monash University, Clayton, Victoria, Australia, 2Observatoire de Physique duGlobe de Clermont-Ferrand, Universite Blaise Pascal, France


The Santorini Archipelago of Greece preserves 650 kyr of volcanic activity, including 12 major explosive eruptions,

which incorporate two 180 kyr mafic to silicic magmatic cycles, as well as the remnants of multiple lava-dome

complexes, lava shields, stratovolcanoes and at least four caldera collapse events. The 3.5 ka Minoan eruption

terminated the second magmatic cycle and was responsible for the destruction of a Minoan settlement at Santorini.

The Pre Lower Pumice 1, Lower Pumice 1 (183.5 ka) and the Lower Pumice 2 (172 ka) eruptions ended the first

magmatic cycle and are considered chemically similar to deposits of the Minoan eruption (dates from Keller et al.

2000). This could suggest cyclicity in magma generation processes.

The Pre Lower Pumice 1, Lower Pumice 1 and Lower Pumice 2 eruption deposits are variably exposed within

the caldera cliffs and outer extremities of southern and eastern Thera. The Pre Lower Pumice 1 sequence is

represented by 13 small volume pyroclastic fallout, ignimbrites and obsidian clast surge deposits, the latter resulting

from the volcanic destruction of a lava dome. A paleosoil separates the Pre Lower Pumice 1 succession from the

overlying pyroclastic fallout (LP1-A), ignimbrite (LP1-B) and lithic-rich lag breccia (LP1-C) deposits of the Lower

Pumice 1 eruption. The Lower Pumice 1 eruption sequence is disconformably overlain by pyroclastic fallout (LP2-A),

ignimbrites (LP2-B), phreatomagmatic ignimbrites with basal layer 1 ground breccias (LP2-C), and lithic-rich lag

breccia (LP2-D) deposits of the Lower Pumice 2 eruption.

Crystal-poor, dacitic white pumice fragments represent the dominant juvenile product within each sequence.

Subordinate abundances of transitional basaltic-dacitic banded and grey pumice fragments, in addition to mafic

scoria, are also present. Chemically, these rock types depict two magma batches: (1) a dacitic (to rhyolitic)

magma, and (2) a basaltic magma. The intrusion and subsequent mixing and mingling of the mafic magma with

a cooling dacitic-rhyolitic magma, is considered responsible for the formation of both banded and grey pumice

fragments within each eruption sequence. C1 chondrite normalised rare earth element (REE) plots, of dacitic white

pumice fragments, depict uniformity in REE patterns within and between eruption sequences. This suggests an

homogeneous source for each eruption and indicates the presence of a long lived magma reservoir. Episodic basalt

magma injection into this reservoir may have triggered each eruption.




Long-term volcanism around active calderas in Bali and Tengger-Bromo region, EastJava, Sunda arc

Kiyoshi Toshida1, Shingo Takeuchi1, Ryuta Furukawa2, Akira Takada2, Supriyati Andreastuti3, NugrahaKartadinata3, Anjar Heriwaseso3, Rosgandika Mulyana3, Asep Nursalim3, Oktory Prambada3, Yudi

Wahyudi3

1CRIEPI, Japan, 2AIST, GSJ, Japan, 3CVGHM, Indonesia


We study the long-term volcanism around the active calderas in Bali and Tengger-Bromo region, East Java.

Mass fractionation correction method is utilised for the mass spectrometry of K-Ar dating, which accounts for the

fractionation of initial argon ratios. Lava samples having pilotaxitic or intergranular groundmass texture are selected

for dating in order to obtain accurate and precise ages. Some of the samples dated are estimated to contain initial

argon ratios that are fractionated from atmospheric values.

We have identified three active periods of volcanism in Bali. They are 1.6-1.5 m.y. BP, 0.7-0.5 m.y. BP, and 0.2

m.y. BP to present. Volcanic rocks to the west of Bratan caldera were formed by the 1.6-1.5Ma activity. Volcanoes

consisting the northern aprons of caldera sommas were formed by the 0.7-0.5 Ma activity. Between 0.2-0.1 Ma,

volcanism occured extensively around present Batur and Bratan calderas. The shield volcanoes consisting the

somma of Batur and Bratan have started to form by covering the 0.5 Ma volcanoes. Batukau, EL 706m volcano

near Pasek, and Cemara volcanoes were also formed in this period. From 0.1Ma to present, the activity continued

at Batur somma and formed Abang peak. Agung volcano started to form by 0.05 m.y. BP. Both Batur and Bratan

systems have produced caldera-forming eruptions multiple times in the past 0.03 m.y., and their intra-caldera activity

has continued along with the activity of Agung.

Caldera-forming eruptions of Tengger-Bromo system are older than Bali calderas, yet the volcano is still very much

active. The two calderas are Ngadisari and Sand Sea. The activity of Tengger volcano started with the formation of

northern somma (Pananjakan) in 0.5-0.45 Ma, after the 1.2 m.y. hiatus following the formation of 1.7 Ma Kukusan

volcano. The aprons were formed in the subsequent active periods. More than half of the edifice volume erupted

between 0.35-0.2 Ma and have formed much of the aprons, with average eruption rate greater than 2km3/ky. The

northeastern to southern somma, and the northern apron, were formed in 0.35-0.3 Ma. The eastern apron and

the southwestern apron were formed in 0.25-0.2 Ma. Lavas and cones in the northwestern apron, and Ngadas

basalt and andesite lavas (which fill the Ngadisari caldera), have formed in 0.08-0.06 Ma. The age of Sand Sea

eruption is determined to be about 0.05 Ma from stratigraphy relations. Therefore, the central cones are found to

be younger than 0.05 Ma, which is consistent to 14C ages of tephra at the caldera rim from previous study. The

average eruption rate of the central cone activity is calculated to be about 0.1 km3/ky. Petrography and whole-rock

chemistry of ejecta from the central cones, including the active vent (Bromo), are similar to those of Ngadas lavas

and Sand Sea eruption PDC deposits. These observations imply that the present magma supply and accumulation

system is similar to the one prior to the Sand Sea eruption.




Vent migration and caldera collapse during the Minoan eruption of Santorini

Timothy H Druitt

Lab Magmas et Volcans, Blaise Pascal University, Clermont-Ferrand, France


The late 17th century BC Minoan eruption of Santorini discharged 30-60 km3 of rhyodacitic magma, and

minor andesitic magma, and collapse deepened and widened a partially flooded 21 ka caldera already

present in the northern half of the volcanic field. The resulting present-day caldera has distinct northern and

southern basins, separated by a NE-SW line of tectonic weakness (Kameni Line, KL). Juvenile and accidental

component populations from the deposits exhibit systematic variations between the four eruptive phases, allowing

reconstruction of vent migration during eruption. The plinian phase (phase 1) discharged rhyodacite and a

suite of compositionally distinctive Ba-rich andesites from a subaerial vent situated on the KL. The vent then

migrated northwards into the ancient caldera, causing eruption of syn-plinian base surges (phase 2) then cohesive,

low-temperature pyroclastic flows (phase 3). The andesitic magma ceased to be erupted as the vent migrated

away from the KL. However, the phase 3 tuffs contain abundant black, glassy andesitic lithics that are chemically

very similar to the juvenile Ba-rich andesite, and which are pieces of a partly subaqueous lava complex present in

the 21 ka caldera prior to the Minoan eruption. Blocks of lava compositionally identical to Minoan rhyodacite were

also discharged in abundance during phase 3. The lithic assemblage is uniform throughout phase 3, suggesting

a single vent situated within the 21 ka caldera, consistent with Pfeiffers (2001) ballistic data. Phase 4 discharged

hot pyroclastic flows that poured into the sea forming three ignimbrite fans (NW, E and S). The first flows (phase

4a) were erupted from one or more vents in the north and laid down lithic-block-rich ignimbrites and lag breccias

containing abundant clasts of Tertiary basement. These constitute much of the NW ignimbrite fan, and contain

gas escape pipes suggesting that some of the lithic debris was wet when incorporated into the flows. Subsequent

flows (phase 4b) were much poorer in block-sized lithics and laid down the hot, fine-grained, tan-coloured ignimbrite

typical of phase 4. Package 4b, which accounts for most of the onland ignimbrite, contains only trace amounts of

basement clasts, makes up the S and E ignimbrite fans, and overlies package 4a in the NW. The greater thickness

of the S and E fans compared to the NW fan suggests that, unlike phase 4a, the phase 4b flows were erupted

from the area of the present-day southern basin. Moreover, their hot, dry nature shows that that the erupting vents

were subaerial, being sited south of the flooded 21 ka caldera. In summary, the principal eruptive vents migrated

from the KL first northwards into the ancient flooded caldera (future northern basin), then southwards onto dry land

(future southern basin). Caldera collapse in the north may have begun as early as phase 3, but the southern half

of the caldera was still subaerial when the eruption ceased.




The 2008 eruption of Okmok volcano, Alaska: Ascent of magma into shallow intracalderagroundwater system and resulting ephemeral landforms

Christina Neal1, Jessica Larsen2, Zhong Lu3, Michael Ort4, Janet Schaefer5

1U.S. Geological Survey Alaska Volcano Observatory, USA, 2University of Alaska Geophysical Institute,USA, 3U.S. Geological Survey Cascades Volcano Observatory, USA, 4Northern Arizona University

School of Earth Sciences and Environmental Sustainability, USA, 5Alaska Division of Geological andGeophysical Surveys, USA


Okmok volcano, a young, largely basaltic andesite caldera system in the Aleutian Islands of Alaska, erupted

explosively over a 5 week period between July 12 and August 23, 2008. The eruption was predominantly

phreatomagmatic, ultimately generating an estimated 0.4 km3 (bulk volume) of tephra that covered much of

northeastern Umnak Island. Eruption onset was abrupt and explosive, producing a tephra column that reached 16

km above sea level in the first four hours. Subsequently, tephra emission was of variable but much lower intensity

through the eruption’s end. Okmok is remote and the eruption was poorly constrained by eyewitness observations.

Four synthetic aperture radar images taken during the eruption, combined with post-eruption fieldwork and analysis

of photography, record a complex sequence and evolution of explosion and collapse craters, and development of a

several hundred meter high tuff cone.

The eruption began as ascending magma encountered shallow groundwater beneath the caldera floor, prompting

multiple explosions along a crudely arcuate zone about 2 km long. At least six overlapping, 100 to 300 m diameter

craters formed within, and adjacent to, a larger crater that had destroyed part of a prominent post-caldera cone.

Continued venting promoted coalescence of the new craters and by day 12 only four distinct, larger craters were

visible. Two of these in the center of the 2 km long zone ultimately formed distinct positive relief features (tephra

cones) that served as primary vents for much of the eruption. Craters at each end of this zone appear to have

hosted sporadic explosions but were not sites of continuous ash emission. These marginal craters grew further in

size through the eruption by additional explosions and collapse. Both filled with water during and after the eruption.

Vigorous post-eruption erosion and deposition are rapidly altering these craters and will likely, within a very short

time, render them unrecognizable as part of the initial 2008 vent system. Similarly, several dozen 10 to 200 m

wide collapse craters that formed north of the primary eruptive vents prominent in the weeks following the eruption

were quickly degraded beyond recognition. The strong involvement of water, complex explosion histories, and

rapid reworking that occurs during protracted phreatomagmatic eruptions may frequently obscure details of initial

vent geometry. Documentation of this process during the 2008 eruption of Okmok volcano, Alaska, serves as a

cautionary note for interpreting older phreatomagmatic deposits and landforms.




Linking lava domes to their structural setting

Paul A Ashwell, Ben M Kennedy, Darren M Gravely, Felix W von Aulock, Jim W Cole

University of Canterbury, New Zealand


The eruption of lava domes can be strongly controlled by local and regional structure, yet it is not always easy

to observe this volcano-tectonic relationship. We infer that the conduits for these eruptions will often be close to,

or along, major faults and so the internal structure of the dome, defined by the shape and position of the conduit,

can also be linked to the structural setting. To investigate the volcano-tectonic relationship, we have focussed our

studies on the internal structures of lava domes that sit within two different calderas in the Taupo Volcanic Zone

(TVZ) of New Zealand. At Ngongotaha Dome, Rotorua Caldera, quarrying has exposed a cross section through

part of the dome, and at Tarawera Volcano, Okataina Caldera Complex, a rift produced during the historical

1886AD eruption exposed part of the interior of the domes. In such a hyperactive caldera-forming region like the

TVZ, it can be difficult to observe the geometry or structure of calderas because of their rapid burial, and thus,

domes may remain as their only structural clue.

At Ngongotaha Dome, a fan-like arrangement of flow bands from the outer edge to the centre suggests a

central dyke-fed conduit running parallel to the outer edge of the dome. The location and morphology of the conduit

can be related to caldera structures, as flow band orientations from within the dome matches elongation of gravity

lows associated with caldera collapse. Post dome growth fractures share similar orientations to the dominant

northeast trending structural grain of the region. We compare the structural orientations within Rotorua Caldera

(gained from mapping of Ngongotaha Dome) to those at nearby Okataina Caldera Complex, in order to investigate

the volcano-tectonic evolution of the regional structure.

Tarawera Volcano is more complex. Previous research has concluded that the domes of Tarawera were

inflated with minor flow development, as suggested by the presence of an onion skin-like circular flow band pattern.

However, flow bands are often near vertical above the proposed conduit, and show a wide variety of orientations

at the flanks of the dome. Instead, a series of vertically emplaced sheets that were then carried outwards from the

vent is suggested as a possible method of emplacement of this dome.

Analysis of the internal structures of lava domes can provide a wealth of knowledge as to the emplacement

of the dome and structural setting it resides within. In particular, where other structural clues like faults are

missing, internal structures and morphologies of lava domes can help to reproduce structural maps at a caldera

scale. These structural maps can also provide insights into the wider regional scale tectonics and evolution of the

structure of the TVZ.




Petrology and geochemistry of the 2011 eruption of Nabro, Eritrea

Amy R Donovan1, Clive Oppenheimer1, Juan Andujar2, Bruno Scaillet2

1Dept of Geography, University of Cambridge, UK, 2Institut des Sciences de la Terre d’Orleans, France


Nabro volcano in Eritrea is a large caldera system that forms a lineament between the Red Sea and the main Afar

spreading system. The largest recorded historical eruption in Africa took place at its neighbour, Dubbi, in 1861

(Wiart and Oppenheimer, 2000). Nabro erupted in 2011, producing a trachy-basaltic lava flow that travelled 15km

along the Ethiopia-Eritrea border, displacing communities and generating the largest SO2 emission since Pinatubo

(Bourassa et al., 2012). Older deposits around Nabro range from trachybasalt to rhyolite in composition, following

a mildly alkaline trend with LREE-enrichment. There is a very large ignimbrite deposit, possibly associated with

caldera formation, around much of the volcano, demonstrating its range of potential impacts. We present new

geochemical and petrological data from the eruption, including whole-rock and mineralogical data. Preliminary

results from experimental studies of the 2011 products will be discussed. These studies provide insights into

tectonic processes in the Afar region and its dynamic evolution, as well as the potential hazards from future

eruptions of a complex caldera system.




Probable time correlation between the eruption of Baitoushan volcano and themegathrust earthquakes in Japan

Hiromitsu Taniguchi

Tohoku University, Japan


From 2002, Baitoushan volcano revealed the sign of magma activity such as the increase in seismicity and

upheaval of the summit. This activity ceased in 2005. After six years from this sign, the 2011 Tohoku earthquake

occurred in Japan. This gave an anxiety of the eruption of Baitoushan volcano as in the case of active volcanoes in

Japan. There is a historical basis in this anxiety like as in 9th century history in East Asia including Japan and the

Asian continent.

To examine the possibility of the Baitoushan eruption, we need the knowledge of historical correlation between

the eruption and megathrust earthquakes in Japan. We have many precise historical records on the megathrust

earthquakes in Japan, contrary to this, we have poor knowledge on the ages of Baitoushan eruption. In this study,

we identified the eruption ages of Baitoushan volcano carefully based on the analysis of old documents. They were

compared with the ages of megathrust earthquakes in Japan.

The eruption age adopted for consideration is following five ages. June 24 1373, October 7-9 1597, June 9 1702,

1898 and 1903. In addition to these ages ca.940 is adopted as the millennium eruption age based on the wiggle

matching dating and the analysis of history of turmoil in Northeast Japan. Although some papers adopted 1403,

1413, 1668, 1900 and 1925 as the eruption age, I did not adopt them because of no description in primary document

or doubt of yellow sand.

Among these eruption ages (6 data), the age interval was 5 to 433 years, and was an average of 193 years. Total

age interval with 3sigma (standard deviation) accuracy was as large as 860 years within the examination period

covered for 1070 years after ca.940. Thus it was meaningless to assume a periodicity. On the other hand, the age

difference (tB-tJ) between the eruption (tB) and the maximum proximity megathrust earthquake (tJ) [4 data sets;

reliable data relatively] ranged from -8 to +12 years, and it was an average of +1.3 years. Within the examination

period covered for 640 years from 1373 with document record to the present, the total age difference with 3sigma

accuracy was only 43 years, enough small to assume the correlation between the eruption and the megathrust

earthquake.

Then, does an eruption occur after the megathrust earthquake of 3.11 in Baitoushan volcano? Supposing it erupts,

when does it occur? There are some facts related on this problem; historical correlation between the Baitoushan

eruption and the megathrust earthquakes along the Japan Trench, recent magma accumulation, and the stress field

change after 3.11 from compressional to extensional. All these facts indicate the possibility of eruption of the near

future. If it will occur in relation to the 3.11 megathrust earthquake, the age will be by 2034 with 3sigma accuracy.




The Mt. Paektu Geoscientific Experiment

Un Young Gun1, Ju Un Ok1, Kim Myong Song1, Ri Gyong Song1, Ri Kyong Nam2, James OSHammond3, Clive Oppenheimer4, Kathy Whaler5, Steve Park6, Graham Dawes5, Kayla Iacovino4

1Earthquake Bureau, Democratic Peoples Republic of Korea, 2State Academy of Science, DemocraticPeoples Republic of Korea, 3Imperial College London, UK, 4University of Cambridge, UK, 5University of

Edinburgh, UK, 6University of California, Riverside, US


Mt. Paektu/Changbaishan volcano is an enigmatic volcano straddling the China — DPRK border. Its most striking

features are the ~5.5–km–wide crater at its summit, and the magnificent lake it hosts. The crater was formed

during one of the largest known historical eruptions, which took place around the middle of the 10th century CE.

The so-called “millennium eruption” resulted in significant tephra fallout as far as northern Japan, highlighting

the potential long-range impacts of volcanism on this scale. Since this event, three smaller eruptions have been

documented, the most recent of which occurred in 1903. More recently, in 2002–2005, elevated levels of seismicity

were recorded beneath the volcano. This has led to increased surveillance and geological work aimed at identifying

the structure of the volcano, and revealing further details of the nature of the “millennium eruption”. We will describe

a joint UK/US/DPRK project engaged in study of the Korean side of Mt. Paektu/Changbaishan. We plan to deploy

a linear array of six broadband seismic stations for one year. Primarily using the receiver function technique, we

will image major crustal discontinuities and attempt to reveal where melt may be ponding in the crust. Further,

we plan to perform audio-frequency magnetotelluric soundings to determine the conductivity of the crust along the

same profile. By highlighting regions of high conductivity and low seismic velocity we will thus show regions of

melt storage beneath the volcano. Additionally, we will investigate both seismic and electrical anisotropy, in order

to infer the geometry of any magma bodies. We will also carry out new work to quantify the nature and impacts of

the “millennium eruption”: this will include experimental studies to determine pre-eruptive conditions, and deposit

characterisation (isopach and isopleth mapping and modelling based on an improved dataset with new borehole

data) for modelling eruptive processes/dynamics. We will also investigate the record of eruptions that have taken

place both prior to and since the “millennium eruption”. We expect to carry out fieldwork in the summer of 2013,

following the meeting in Kagoshima and also in summer 2014. This contribution will therefore focus on reporting on

the goals and planning for this international project.




Formation of Qixiangzhan Eruption of Changbaishan Tianchi Volcano, China

Bo Pan1, Jiandong Xu1, Guangpei Zhong2, Guoming Liu2, Bo Zhao1

1Institute of Geology, China Earthquake Administration, China, 2Changbaishan Volcano Observatory,Jilin, China


To comprehensively understand the eruption characteristics and history of active volcano is crucial for predicting

its future eruptions and hazard. Changbaishan Tianchi Volcano (CHTV) is one of the most dangerous active

volcanoes in Northeast Asia. It had experienced three periods of large-scale eruptions since the Holocene, i.e.

the Tianwenfeng period at about 5,000 years ago, the Qixiangzhan (QXZ) period at about 4,000 years ago and

the Millennium eruption at about 1,100 years ago, respectively. The type of Tianwenfeng and Millennium eruptions

is commonly accepted to be a typical Plinian eruption. However, there arises a considerable debate about the

type of QXZ eruption as to whether it is effusive or explosive. In high-resolution remote-sensing images, the

morphology of the products of QXZ eruption looks like a lava flow, which flows along the northern slope of the

volcanic cone about 5.4 km in length and 400-800 m in width. However, the recent research work by the author

has revealed that the QXZ eruption should be a small-scale pulsed explosive eruption. The main evidence is as

follows: 1) The bulk-rock composition of QXZ eruption products is characterized by high SiO2>71% and Na2O+K2O

>10% contents representative of alkaline magma, which has high viscosity, low flowing ability and extremely high

potential of explosive eruption; 2) Field observations show that the QXZ eruption products appear as thin layers

about 2-5cm in thickness, significantly different from the massive or slaggy structures of lava flow; 3) Microscopic

observation reveals that most of the phenocrysts in the QXZ eruption products were severely broken by explosion

to form angular grains with well developed micro-cracks. The vesicles in the QXZ eruption products are irregular

in shape and have rough margin, different significantly from the elliptical and smooth margin vesicles commonly

observed in lava flow; 4) Stereomicroscopic observation shows that the QXZ eruption products are composed of

clastic particles and exhibit grain-supported texture with well developed irregular vesicles. Further study by using

SEM indicates that most of the clastic particles are fine volcanic ash and tiny pumice with rough surfaces. Basing

on the above analyses, we may conclude that the QXZ eruption can be assigned to a small-scale pulsed explosive

eruption. During the explosive eruption, a large number of fine pyroclastic particles flowed down the mountain

slope as a high speed pyroclasstic flow to form thin layer of ignimbrite. Over many times of explosive eruptions,

layer upon layer of ignimbrite were accumulated, resulting in a shape just like lava flow. Therefore, all the three

large eruptions of CHTV in Holocene can be assigned to explosive eruption, rather than the previously proposed

model of explosive-effusive-explosive explosions.




Textural characteristics of the Holocene pumice erupted from Changbaishan volcano andtheir volcanological implications

Hongmei Yu1, Jiandong Xu1, Jianping Wu2, Chuanyong Lin1

1Institute of Geology, China Earthquake Administration, China, 2Institute of Geophysics, ChinaEarthquake Administration, China


The texture of volcanic products records abundant information about the physical and chemical processes of the

magma. Therefore, the quantitative analysis of the texture of volcanic products has recently become an important

research method in volcanology. Changbaishan volcano located on the border between China and North Korea, is

one of the largest active volcanoes in China. It has experienced at least five explosive eruptions in the Holocene,

i.e. 5,000 BP eruption, AD 946 eruption (also called "millennium eruption"), AD 1668 eruption, AD 1702 eruption

and AD 1903 eruption. In this study, the composition and quantitative texture of pumices from three explosive

eruptions in Holocene (5000 BP eruption, the millennium eruption and AD 1668 eruption) of Changbaishan volcano

were studied in detail. The results show that, the pumices from 5000 BP eruption and the millennium eruption

are all pantellerite in composition, but the later is more acid than the former. The pumices from 1668 AD eruption

are high potassium trachyte in composition. The pumices from these three eruptions comprise mainly vesicle of

different sizes, vesicle wall and a small amount of phenocrysts ( <15% ). The parameters of vesicles are extracted

from back scattering SEM images with different magnifications. The pumice from 5000 BP eruption has the highest

vesicularity (about 73.4%), the smallest size (about 1 um) and the greatest number density (4.23 ∗ 1016 m -³) of

vesicles. The pumice from the millennium eruption has the vesicularity of 62.8%. The smallest vesicles in this

pumice are several micrometers in size, and the number density is 3.25 ∗ 1015 m -³. The pumice from AD 1668

eruption has the vesicularity of only 45.9%, the vesicles are generally larger than 10 um in size, and the largest

may reach up to 1 cm. In pumices of this period, the number density of vesicles (3.68 ∗ 1014 m -³) decreases and

the vesicle walls become thicker. Finally, according to the compositions and number density of vesicles in pumices

from three eruption periods, we have estimated some important physical parameters, such as the decompression

rates, height of eruption column and magma discharge rate for these three eruptions. The results from this study

may provide important scientific basis for understanding magma process and determining the intensity of historical

eruptions of Changbaishan volcano.




Characteristics of Pyroclastic-flow Facies in Millennium Eruptionin of TianchiVolcano,Changbai Mountains,China

Bo Zhao, Jian Dong Xu, Bo Pan, Mei Hong Yu

Institute of Geology,China Earthquake Administration, China


Tianchi volcano is the largest composite volcano in China, which is at the boundary between China and North

Korea. In 946±3, there was a large plinian eruption in Tianchi volcano which is called millennium eruption.

Millennium eruption produced widespread pyroclastic-flow deposits.

This paper has studied thestrata, litho, geochemistry, grain-size and transportation of ground-surge from

pyroclastic-flow of millennium eruption in Tianchi volcano. The main conclusions which have been got are as

follows:

(1) The proximal strata of pyroclastic-flow are combined with the eutaxite structures and lava-like structures. The

median strata are combined with massive beds, pumice-rich layers, litchis-richer layers and welded zones which

are representative of the transportation processes for gravitational differentiation. And columnar joints and block

structures are also developed which are representative of the deposition processes for cooling. The distal strata are

combined with coarse-tail beds, ground surge beds and climbing beds. There are also some altered beds because

of flood. The distal pyroclastic-flow strata have some fluidized characteristics.

(2) The proximal ignimbrite is alkaline trachytic welded tuff. The median ignimbrite is also alkaline trachytic welded

tuff. And the distal ignimbrite is alkaline rhyolite. The welded tuff is weaker from the proximal to the distal.

(3) The proximal and median black pumice is trachyte. And the distal gray pumice is rhyolite.

(4) The median diameters of pumice which is less than 64mm from the median and distal strata have an increasing

tendency with the distance increasing from the crater. And the median diameters become smaller with the depth

increasing from the top of strata. The Sphericity (Spht) has an increasing tendency with the distance increasing

from the crater. And the contents of lithoclasts in number percentage decrease with the distance increasing from

the crater which reveals gravitational differentiation.

(5) With the pumice becoming smaller, there are more angles, richer irregular shapes and simpler transportation.

The grain-size distributions of fine ashes are similar and have a single peak which is close to the fines.

(6)The histograms of ashes cloud and ground surge have a similar characteristic which is ladder-like and close

to the fines. And the histograms of pyroclastic-flow have many peaks which are representative of the composite

transportation.

(7) There was hydration at the distal part of pyroclastic-flow, and the ground surge came from pyroclastic-flow and

water interaction in Baixi.




History of Volcanic Activity, Magma Evolution and Eruptive Mechanisms of theChangbaishan Volcanic Province

Qicheng Fan, Jianli Sui, Ni Li, Yongwei Zhao

Institute of Geology, China Earthquake Administration, China


Three giant stratovolcano near the border of Sino-DPRK , including Nanbaotai volcano (2343 meters in altitude),

Wangtiane volcano (2051 meters in altitude) and Tianchi volcano (2749 meters in altitude), record the migration

path of volcanism in this area. All the volcanoes endured shield-forming and cone-forming stages. Wangtian‘e

volcano began to erupt in Pliocene and fall into silence in the early Pleistocene (4.77 Ma-2.12 Ma) (Fan et

al.,1998). Tianchi volcano endured three eruption stages: potassium trachybasalt shield-forming stage in the early

Pleistocene (2.77-1.203Ma), trachyte cone-forming stage in the middle and late Pleistocene (1.12-0.04Ma) and

violent eruption of pantellerite in the Holocene (Fan et al., 2006, 2007).

The history record has implied many eruptions in the Changbaishan area in the Holocene time, whereas it is

still hard to identify whether the exact eruption spot is in Tianchi volcano. Although scholars reach an accord

on the Millennium eruption events in Tianchi volcano, the vague history records and the decoupled geological

dating data are still under debating. On the south and north slope of Tianchi volcano, white pantellerite pumice are

covered by black melted trachyte pyroclastic rocks, indicating a medium eruption has happened after the Millennium

eruption events. Studies on major and trace elements and isotope geochemistry show that the Changbaishan

volcanoes have a similar trachybasaltic magma system with its primary magma of potassic trachybasalt nature.

The composition of the magma experienced a evolution from basaltic to trachytic to pantellerite (Fan et al., 2006,

2007).

The studies suggest that there are crustal magma chamber and mantle magma chamber under the Tianchi

volcano. The two chambers interact with each other and cause batch eruption (Fan et al., 2007). Since the mantle

chamber continually supplied trachybaslt influx from below to the crsutal chamber, the Tianchi Volcano is a long-life

volcano. Fractional crystallization and magma mixing were two important processes in the Tianchi Volcano-the

former determined the bimodal characteristics of the volcanic rocks, and the latter triggered the volcanic eruption.

Additionally, some mantle sourced magma of potassic trachybasalt compostions erupted directly to the surface and

resulted in widespread isolated small cinder cones around the Tianchi volcano.

Subduction of the western Pacific plate and the subsequent back-arc extension of NE Asia dominate the mechanism

of the Changbaishan Volcano.





ORIGIN OF CENOZOIC BASALTIC LAVAS IN THE ERDAOBAIHE RIVER VALLY,CHANGBAISHAN, CHINA

Qingfu Yang1, Xiang Liu2, Fengxue Chen1, Ruoxin Liu3

1Earthquake Administration of Jilin Province, Changchun, China, 2College of Earth Sciences, JilinUniversity, Changchun, China, 3Institute of Geology, China Earthquake Administration, Beijing, China


The Tianchi volcano in the Changbaishan on the border between China and North Korea is composed of the

Pliocene to early Pleistocene basaltic shield, middle to late Pleistocene trachytic composite cone. It is an active

volcano because of several eruptions in the Holocene.

In this study, we surveyed basaltic lava sections in the Erdaobaihe river valley and analyzed chemical composition

of the lavas. They consist of basalt in upper level and trachybasalt and basaltic trachyandesite in middle to

lower levels. The lavas are similar in REE patterns, suggesting that they likely share the same source. In the

diagram of Na2O-K2O and La-La/Yb,basalts are shown as Na-rich and tholeiite series, whereas, trachybasalt

and basaltic trachyandesite are K-rich and alkali-series.Their Mg#(=100Mg2+/(Mg2++Fe2+)<60) are lower than that

(=60-68) of Cenozoic primary basalt magmas in the eastern China. Their K/Rb ratios (0.05-23.15) and Ni contents

(27.76-200.61) are less than primary mantle, whereas Ba/Rb ratios(15.64-264) are greater than primary mantle. All

lines of evidence indicate that these lavas originate from evolved magma. The SI values for basalt and trachybasalt

are less than 40, and SI values for basaltic trachyandesite are in the range of 20−29. The lavas contain phenocryst

of olivine, pyroxene and plagioclase. There are positive anomaly of Ba in trace element pattern and negative

anomaly of Eu in REE pattern. These geochemical features reveal that tholeiite formed from mantle derived magma

which underwent crystal fractionation. Higher K2O (>2.0%) of trachybasalt and basaltic trachyandesite indicate that

variable degrees of contamination may occur during magmas ascending. The discrimination diagrams of Zr-Nb-Y

and Zr/Y-Zr show that the lavas formed in inner plate setting.

This study was funded by CEA(201208005) and NSFC (41072249).




Volcanic geology, geochronology and geochemistry information provided by the HalahaRiver and Chaoer River volcanic field in Daxing’an Mountain Range

YongWei ZHAO, QiCheng FAN

Institute of Geology, China Earthquake Administration, China


34 Quaternary volcanoes, which scattered along a Quaternary NE strike fault, are found in the area of Halaha River

and Chaoer River, middle of Daxing’an Mountain Range. Quaternary volcanic rocks in this area, mainly alkaline

basalt, cover an area of ca. 400 km2. Both magmatic eruptions and phreatomagmatic eruptions are found in this

area. The magmatic eruptions form a series of products, including scoria cones, tephra fallout deposits, and many

kinds of lava (e.g. aa lava, pahoehoe lava and block lava). Typical fumarolic cones and lava hillocks are found in

the lava. The phreatomagmatic eruptions have typical base surge deposits, which are characterized by parallel

bedding and staggered bedding. Volcanic activity in this area form many volcanogenic lakes. According to the

difference of lake forming, they are divided into four types: crater lake, maar lake, volcanic dammed lake, collapse

lava lake (Zhao et al., 2008).

Based on studies on the volcanic field characteristics, in conjunction with geological dating by K-Ar, it is identified

that the volcanism occurred in four periods: Early Pleistocene, Middle Pleistocene, Late Pleistocene and Holocene.

Basalts of Early Pleistocene, mostly mantled by the later volcanic rocks, are distributed in the margin and valleys

of the volcanic field. Middle Pleistocene, the most volcanic active period in this area, witnessed the formation

of more than half of Quaternary volcanoes and lava spreading. Moderate volcanism occurred in Late Pleistocene

which produced a small amount of volcanic deposits. Volcanic activities are strengthened again in Holocene Period,

characterized by strongly explosive explosion, widespread lava flow and well-keeping lava landforms features (Fan

et al., 2011).

The volcanic rocks, dominated by alkali olivine basalts in sodium series, is characterized by relative enrichment in

large ion lithophile elements and light rare earth elements. The fractionation of rare earth element of the basalts is

weak((La/Yb)N =8-12).They resemble alkali basalts in Datong, as shown by trace elements distribution patterns, and

generally exhibit OIB-like characteristics. The basalts show nearly homogeneous Sr-Nd-Pb isotopic composition

similar to MORB source and present depleted mantle characteristics. All data show that basalts of HC have a

garnet lherzolite mantle source, low degree partial melting(8%-15%)in which results in the primitive magma. Crystal

fractionation of olivine and pyroxene from the magma is weak and seldom contamination by the crust rocks happens

during the magma ascending, which resulting the volcanic rocks with high MgO content(>9wt%), Ni content and

Mg value(60-70).Regional extension triggers asthenospheric upwelling, which may lead to the genesis of magma

and subsequent volcanism (Zhao and Fan, 2012).





Origin of Cenozoic basaltic magmatism at Changbaishan volcanic field, northeast China

Takeshi Kuritani1, Jun-Ichi Kimura2, Eiji Ohtani3

1Osaka City University, Japan, 2Japan Agency for Marine-Earth Science and Technology, Japan,3Tohoku University, Japan


In NE China, Cenozoic intraplate basalts are widely distributed, and Changbaishan is one of the largest volcanic

fields. This intraplate magmatism is unique because the stagnant slab of the subducted Pacific plate is present

in the mantle transition zone (MTZ). Therefore, the geodynamic and geochemical processes in relation to the

stagnant slab and their effects on the magmatism have received considerable attention (e.g., Zou et al., 2008;

Ohtani and Zhao, 2009). This paper summarizes the results of our recent studies on the Cenozoic basalts in

NE China (Kuritani et al., 2009, 2011, 2013), and integrates them toward understanding the origin of the basaltic

magmatism at Changbaishan.

A prominent low-velocity anomaly with a plume-like shape was imaged in the upper mantle beneath Changbaishan

by P-wave tomography, and was suggested as an upwelling of wet materials from the MTZ (Zhao et al., 2009). We

have compiled geochemical data of Quaternary basalts from Changbaishan and the surrounding volcanic fields

(e.g., Jingbohu and Kuandian), and found a spatial correlation between the basalt chemistry and the distribution of

the low-velocity zone. Namely, the Ba/Th and Pb/U ratios of the basalts tend to decrease with increasing distance

from Changbaishan. Furthermore, the basalts from Changbaishan have Sr, Nd, and Pb isotopic ratios similar to

the EM1 endmember and the rests are from more depleted source.

One possible source of the EM1-like component is the lower part of the sub-continental lithospheric mantle (SCLM).

However, this is unlikely because the lower part of the SCLM beneath NE China formed in Cenozoic times,

precluding prolonged isotopic ingrowth. Alternatively, the component was proposed to be derived from the MTZ.

In the subducted materials, K-hollandite in metamorphosed sediments can be a major reservoir of incompatible

elements such as Sr, Ba, and Pb. Elements released by breakdown of sedimentary K-hollandite would be

high-Ba/Th and Pb/U, and thus, a suitable metasomatic agent to form EM1-like mantle in the MTZ (Rapp et al.,

2008). Because the formation of the EM1 isotopic signature requires a timescales of >1 Ga, sediments in an

ancient stagnant slab, as well as those in the Pacific plate stagnant slab, may have been involved in the source

materials for the Changbaishan basalts.

Assuming that the age of the ancient slab sediment is 1.5 Ga, the isotopic ratios of the Changbaishan basalts can

be explained by mixing of a depleted mantle with a 0.5% sediment component which consists of a 1:2 to 1:3 mixture

of the recent and the ancient sediments. The MTZ beneath NE China is remarkably hydrous, as evidenced from

the electrical conductivity observations (Kelbert et al., 2009). Therefore, it is plausible that a hydrous mantle plume

ascended from the hydrated MTZ, and that partial melting occurred in the ascending plume in the asthenospheric

mantle, leading to the basaltic magmatism at Changbaishan and its surrounding volcanic fields.




New insights from pumice-rich submarine density currents from caldera-formingeruptions in the Izu Bonin arc (ODP 126)

Martin Jutzeler, James DL White

University of Otago, New Zealand


The Sumisu rift in the Izu Bonin arc was the focus of numerous studies in the late 1980s. During the ODP expedition

126, the upper arc was piston-cored to depths down to 250 mbsf. Most research was focussed on the forearc, with

few investigations addressing the enormous volume of volcaniclastic deposits at sites 790 and 791 in the rift basin,

and site 788 on the rift shoulder. At site 790C, the piston cores reached 275 ka at 180 mbsf, and sampled numerous

units of silicic pumice lapilli developed in beds up to several 10s of m thick, and interbedded with units of silicic and

basaltic ash and hemipelagic mud. Careful analysis of the stratigraphy from core to core allows reconstruction of

the major eruption sequences, although many of the very thick pumiceous beds were partially disturbed during

core recovery, artificially increasing their thickness, and creating false bed boundaries. From our new logging data

from legacy samples stored at the Kochi Core Center, we present a first attempt at stratigraphic correlation of beds

throughout the rift basin, and discuss the origin of the pumice lapilli beds. Textural analyses, including of grain size,

grain shape, componentry and pumice vesicularity were carried out to characterise the most recent pumice lapilli

beds, in addition to clast, glass and mineral geochemistry. This new dataset allows correlation of strata across

the rift basin with high confidence, underpinning our analysis of the style and environment of eruption, as well as

of transport processes, involved in producing the pumice lapilli beds. We expect the pumice lapilli beds to have

been derived from caldera-forming eruptions and by post-eruptive resedimentation, and discuss the probable water

depth at the vent during eruption. This work applies modern textural and geochemical studies to legacy cores,

showing the long-term value (already 20 yrs after collection) of samples from coring the shallow crust.




A thermomechanical perspective on caldera …models to Santorini Volcano in Greece where recent...

Documents

Transcript of A thermomechanical perspective on caldera …models to Santorini Volcano in Greece where recent...